Xiao Hu

Interests

No activities entered.

Courses

Scholarly Contributions

Chapters

• Dong Ng, J. T., & Hu, X. (2024). Digitising cultural heritage in Hong Kong: An overview. In Digital Humanities and Intelligent Computing of Cultural Heritage: Global Development and China Solutions. Taylor and Francis. doi:10.4324/9781032707211-9
  More info

  As a special administrative region in China with a unique historical root, Hong Kong stands as an international financial centre and a multicultural metropolitan city. Rich in not only economic but also cultural resources, Hong Kong houses a remarkable number of statutorily graded historic buildings, officially recognised intangible cultural heritage items, and movable cultural heritage to be systematically documented. A variety of stakeholders in Hong Kong have initiated, supported, and participated in different types of cultural heritage digitisation projects. Nevertheless, there is a lack of a comprehensive and up-to-date overview of these projects. To fill this gap, this chapter surveys cultural heritage digitisation projects in Hong Kong both in the literature and in practice. In particular, this chapter summarises the approaches adopted in these projects and reports the development trends. Our findings reveal the multi-layered objectives of these projects, their different funding sources and modes of collaboration, their selection of cultural heritage types and project themes, as well as their adoption of technologies and user involvement. Based on these findings, this chapter discusses the multifaceted facilitating factors, limitations, and challenges of these projects, and recommends best practices for future projects in Hong Kong and beyond.
• Koh, E., & Hu, X. (2023). Learning analytics for learning: Emerging international trends and case studies from the Asia-Pacific. In International Handbook on Education Development in the Asia-Pacific. Springer Nature. doi:10.1007/978-981-19-6887-7_54
  More info

  In this chapter, an overview of learning analytics is provided-highlighting emerging international trends-illustrated with innovative case studies from the Asia-Pacific. As a growing field intersecting learning and pedagogical theories, human-centered design, and data science, learning analytics has many applications in K-12 and adult learning settings, from enhancing learning progress and learning awareness, improving cognitive learning outcomes, nurturing socioemotional and lifelong learning skills, to intervening with prompts, tasks, feedback, and learning strategies. While there are many recent movements such as multimodal learning analytics, trustable data, and actionable dashboards, they essentially drive towards the ultimate purpose-learning analytics is for optimizing learning. Upon reviewing influential literature in the field, we conceptualize a framework to map current research trends in learning analytics into seven dimensions, including the foundational lens, visual feedback, indicators and metrics, design approach, function/purpose type, data modality, and ethics. This framework demonstrates a global convergence in the field with wide application including the Asia-Pacific region. Case studies of learning analytics applications from Hong Kong and Singapore are illustrated to highlight the fruitful ways how learning and learning environments have been optimized, along the dimensions in the framework. The chapter concludes with a synthesis and critique of current learning analytics research and suggests implications for learning analytics researchers, developers, and users including practitioners.
• Hu, X., Lei, L. C., Li, J., Iseli-Chan, N., Siu, F. L., & Chu, S. K. (2016). Access moodle using mobile phones: Student usage and perceptions. In Mobile Learning Design: Theories and Application. Springer International Publishing. doi:10.1007/978-981-10-0027-0_10
  More info

  This study investigated how often students used mobile phone to access various activities on Moodle. A survey on self-reported usage was filled by 252 university students in courses offered by four different faculties at the University of Hong Kong. Follow-up interviews were conducted to solicit students’ perceptions on mobile access to Moodle and the underlying reasons. The results show significant differences in students’ usage of various Moodle activities via mobile phones. Students’ responses also suggest that mobile access to Moodle is a necessary complement to computer access but its limitation on usability and reliability may have restricted its potential in enhancing teaching and learning.

Journals/Publications

• Ba, S., & Hu, X. (2025). Effects of background music tempo and mode on reading comprehension: the mediating role of emotions. Instructional Science, 53(Issue). doi:10.1007/s11251-025-09728-5
  More info

  It is common for learners to listen to background music during learning activities. However, existing research has produced inconclusive results on the effects of background music on learning due to different circumstances and contexts. There is a lack of detailed understanding regarding the effects of various background music properties on learning. Based on the arousal-and-mood hypothesis, this study investigates how background music with different tempi (fast and slow) and modes (major and minor) affects reading comprehension through the mediation of learning emotions. A within-subject research was conducted with 33 graduate students who listened to five pieces of background audio while completing reading comprehension tasks. Particularly, a multimodal approach was adopted to capture participants’ emotions through self-report scales, facial expression recognition, and physiological monitoring. Using a series of linear mixed models for data analysis, results demonstrated that background music with different tempi and modes were associated with significantly different self-report emotion measures, facial expressions, and physiological responses. Moreover, while music properties did not directly predict reading comprehension performance, music-induced emotions measured by self-reported valence, body temperature, and heart rate were significantly associated with reading comprehension performance. Therefore, this study establishes through multimodal emotion measures that background music properties have indirect effects on reading comprehension performance through the mediation of music-induced emotions. This study refines the current understanding of background music and contributes to the music selection and design for learning.
• Hu, X., & Ng, J. T. (2025). Low tech-barrier virtual reality content creation for cultural heritage education: learning outcomes and pedagogy. Technology, Pedagogy and Education. doi:10.1080/1475939x.2025.2503768
  More info

  Aligned with the United Nation’s Sustainable Development Goals, virtual reality (VR) content creation helps novice learners better understand about cultural heritage, yet its pedagogical potential remains under-explored. Synthesising the literature, this study developed a multidimensional framework of learning outcomes in the interdisciplinary domain of cultural heritage education. Based on the principles of designing maker activities, this study designed a novel, research-informed and low tech-barrier pedagogical approach of VR content creation. Applying it to a general education course on digitising cultural heritage, the authors analysed instructor-given, rubric-referenced performance scores and qualitatively examined reflections from 302 undergraduate students of diverse disciplinary backgrounds. Results have shown that the maker activity of VR content creation outperformed other traditional learning approaches in terms of students’ learning outcome achievement. Based on empirical evidence, this study verified the learning outcome framework, yielded the corresponding pedagogical implications and laid the groundwork for future studies on VR-based maker activities.
• Law, N., Wang, N., Ma, M., Liu, Z., Lei, L., Feng, S., Hu, X., & Tsao, J. (2025). The role of generative AI in collaborative problem-solving of authentic challenges. British Journal of Educational Technology. doi:10.1111/bjet.70010
  More info

  This study investigates undergraduate and postgraduate teamwork in a four-week ‘Generative AI for Social Good’ hackathon, focusing on how students use GAI tools in authentic problem-solving within their learning ecology. It examines the factors that foster productive collaboration and explores evidence of AI extending human cognition beyond mere tool use. Data sources included pre- and post-surveys, interim reports, submitted artefacts and team workspace logs. Generative AI (GAI) use accounted for nearly half of the demonstrated digital competence instances—particularly in content creation and problem-solving—highlighting its role in facilitating collaborative, inquiry-driven learning. Findings reveal that success depended not on computational expertise, but on shared values, diverse skill sets, effective team structures and clear communication. GAI's role evolved with teams' technical backgrounds, dynamically supporting collaborative knowledge building and moving beyond instrumental use to actively shape emerging knowledge building. These insights offer valuable implications for the pedagogical design of learning with and through GAI.
• Liu, Y., Hu, X., Ng, J. T., Ma, Z., & Lai, X. (2025). Ready or not? Investigating in-service teachers’ integration of learning analytics dashboard for assessing students’ collaborative problem solving in K–12 classrooms. Education and Information Technologies, 30(Issue 2). doi:10.1007/s10639-024-12842-5
  More info

  Collaborative problem solving (CPS) has emerged as a crucial 21st century competence that benefits students’ studies, future careers, and general well-being, prevailing across disciplines and learning approaches. Given the complex and dynamic nature of CPS, teacher-facing learning analytics dashboards (LADs) have increasingly been adopted to support teachers’ CPS assessments by analysing and visualising various dimensions of students’ CPS. However, there is limited research investigating K-12 teachers’ integration of LADs for CPS assessments in authentic classrooms. In this study, a LAD was implemented to assist K-12 teachers in assessing students’ CPS skills in an educational game. Based on the person-environment fit theory, this study aimed to (1) examine the extent to which teachers’ environmental and personal factors influence LAD usage intention and behaviour and (2) identify personal factors mediating the relationships between environmental factors and LAD usage intention and behaviour. Survey data of 300 in-service teachers from ten Chinese K-12 schools were collected and analysed using partial least squares structural equation modelling (PLS-SEM). Results indicated that our proposed model showed strong in-sample explanatory power and out-of-sample predictive capability. Additionally, subjective norms affected technological pedagogical content knowledge (TPACK) and self-efficacy, while school support affected technostress and self-efficacy. Moreover, subjective norms, technostress, and self-efficacy predicted behavioural intention, while school support, TPACK, and behavioural intention predicted actual behaviour. As for mediation effects, school support indirectly affected behavioural intention through self-efficacy, while subjective norms indirectly affected behavioural intention through self-efficacy and affected actual behaviour through TPACK. This study makes theoretical, methodological, and practical contributions to technology integration in general and LAD implementation in particular.
• Omeh, C. B., Olelewe, C. J., & Hu, X. (2025). Application of artificial intelligence (AI) technology in tvet education: Ethical issues and policy implementation. Education and Information Technologies, 30(Issue 5). doi:10.1007/s10639-024-13018-x
  More info

  The adoption of generative AI in educational process carries both potential advantages and risks hence there is a need for ethical principles to guide its adoption in education. A population of 443 TVET educators, including 325 male and 118 female, was selected for this study using a mixed research design from the seven TVET public institutions in south-eastern Nigeria. The study was guided by three research questions and three null hypotheses. The instruments used for data collection were a structured questionnaire and a guided interview developed by the researchers in line with the research questions. The Cronbach Alpha reliability test, which produced a reliability index of 0.9, was used to determine the internal consistency. Mean, standard deviation, factor loading and t-test were used to evaluate the data, and an independent t-test with a significance level of 0.05 was used to test the null hypotheses. The findings of the study indicated that TVET educators agreed that AI technology is an effective educational technology, there is a need for ethical principles to ensure data privacy, data integrity, data reliability, data transparency and data accuracy are not compromised. Also, TVET educators agreed that the adoption of AI technology in the educational process improves academic performance, increases learning engagement, and supports classroom inclusion and personalized learning. It was recommended that tertiary institutions broaden their current AI policies in light of the findings of this study to support the successful integration of AI technology into the educational process.
• Que, Y., & Hu, X. (2025). Enhancing Learners' Reading Comprehension With Preferred Background Music: An Eye-Tracking, EEG, and Heart Rate Study. Reading Research Quarterly, 60(Issue 2). doi:10.1002/rrq.70004
  More info

  It is often observed that many learners prefer reading texts with background music (BGM), which may enhance engagement and mood in reading. However, the impact of BGM on reading comprehension performance or process remains heterogeneous. This study examined how learners' self-provided preferred BGM affected reading comprehension accuracy and cognitive processes as measured by self-reports and multimodal psychophysiological signals (e.g., eye movements, EEG signals, heart rates) during reading, while also considering the role of personal traits and music characteristics. Data were collected from a within-subject experiment with 52 nonnative English speakers, who read half of the English text passages with their self-provided BGM and the other half in silence. The results indicated that overall BGM did not significantly affect reading task performance, self-reports, or multimodal psychophysiological responses. Hierarchical linear modeling revealed that learners' personal traits (e.g., working memory capacity, multitasking ability, language proficiency, BGM listening habits, prior knowledge) moderated the effects of BGM on their cognitive processes during reading as indicated by the psychophysiological responses. Stepwise linear regression showed that the presence of lyrics predicted decreased reading comprehension accuracy and disrupted word-level lexical processing, while a fast tempo was related to more efficient text processing. These findings offer empirical evidence on how BGM influences learners' reading task performance and cognitive processes and provide practical guidance for stakeholders (e.g., learners, instructors, learning environment designers) on the personalized and appropriate use of BGM for enhancing reading comprehension.
• Que, Y., & Hu, X. (2025). Examining the Effect of Background Music on Learners’ Attention and Cognition in Virtual Reality Environments: A Psychophysiological Study. International Journal of Human-Computer Interaction. doi:10.1080/10447318.2025.2505778
  More info

  Virtual Reality (VR) enriches learning and instruction, while background music (BGM) is widely employed to modulate attention and cognition. Understanding how BGM influences learning in VR is essential for optimizing VR learning environments and has the potential to improve educational outcomes, yet this area remains largely unexplored. This study collected fifty-two participants’ self-reports, electroencephalogram signals, eye movements, heart rates, and interview responses to explore their attention and cognition during studying virtual heritage sites with and without BGM, while considering individual traits as influential factors. Results showed that with BGM, participants reported higher levels of engagement and demonstrated longer fixation duration on heritage sites’ image regions, than without BGM. Moreover, participants’ self-reported familiarity with the cultural heritage site and BGM listening frequency moderated the effect of BGM on attention and cognition in VR. This study offers implications for incorporating BGM for learning in VR, and designing personalized VR learning environments.
• Que, Y., & Hu, X. (2025). Exploring background music for studying in virtual reality learning environments. Interactive Learning Environments, 33(Issue). doi:10.1080/10494820.2025.2483412
  More info

  Researchers have delved into the advantages and obstacles associated with incorporating background music (BGM) into learning and instruction, but the influence of BGM in educational virtual environments remains understudied. In addressing this research gap, this study leveraged virtual reality (VR) technology and developed a novel setting for informal learning, where 52 students engaged with eight immersive virtual cultural heritage sites. The study followed a within-subjects design, with each student experiencing half of these sites with BGM, and the other half in silence. We examined how learners’ task performance, emotional changes, and perceptions (of immersiveness and enjoyment)–assessed through comprehension tests, self-reported questionnaires, and interviews–differed between BGM and silence conditions, and how they were related to BGM characteristics and learner traits. Results revealed that, while studying cultural heritage in the VR environment, (1) BGM enhanced learners’ enjoyment and sense of immersiveness, without compromising their comprehension accuracy and (2) BGM characteristics and learner traits significantly correlated with learners’ performance, emotional changes, and perceptions. Findings further our understanding of how BGM affects learning and provide design implications for leveraging BGM to enhance learning in VR environments. Future research could explore how BGM affects diverse VR learning contexts.
• Que, Y., Zheng, Y., Hsiao, J. H., & Hu, X. (2025). Using eye movements, electrodermal activities, and heart rates to predict different types of cognitive load during reading with background music. Scientific Reports, 15(Issue 1). doi:10.1038/s41598-025-03052-1
  More info

  The triarchic model of cognitive load postulates three types of cognitive load—extraneous, intrinsic, and germane load. While various approaches have been proposed to measure the three types of cognitive load, most measurements are intrusive. To address this issue, we leveraged multimodal learning analytics to collect eye movement (EM), electrodermal activity (EDA), heart rate (HR), and heart rate variability (HRV) from non-intrusive sensors and investigate whether they could predict the three types of cognitive load. We examined extraneous load (created by adding background music (BGM)), intrinsic load (created by text complexity), and germane load (reflected by comprehension accuracy) in a novel reading context with self-selected preferred BGM. One hundred and two (102) non-native English speakers were recruited. Half of them read English passages with BGM, while the other half read in silence. Results of logistic regression indicated that EM measures were predictive of the three load types, while HR/HRV measures predicted extraneous and germane load. Our findings provide evidence supporting the triarchic structure of cognitive load theory and implications for the design of non-intrusive measurement of cognitive load.
• Su, J., Chen, X., Chu, S. K., & Hu, X. (2025). A scoping review of empirical research on AI literacy assessments. Educational Technology Research and Development. doi:10.1007/s11423-025-10515-9
  More info

  AI literacy is becoming increasingly popular in education, yet there has been limited focus on reviewing empirical research on AI literacy assessment. The purpose of this study was to synthesize existing empirical studies on AI literacy assessment, with the aim to understand how AI literacy has been assessed and to inform future AI literacy assessment development. This scoping review evaluates and synthesizes 36 studies on AI literacy assessment published between 2019 and 2024, involving assessment tools, forms of assessment, and reliability and validity evidence related to AI literacy assessment. The review proposes four aspects (i.e., knowledge of AI, AI ethics, affect towards AI, and use of AI) for assessing AI literacy. The results showed that (1) most research focused on assessing primary and secondary school students’ AI literacy; (2) most studies used questionnaires, followed by surveys; (3) most studies used computer-based tests, followed by paper-based tests; (4) most studies assessed participants’ AI knowledge, followed by AI ethics; and (5) only a few studies reported evidence of the reliability and effectiveness of their assessments. Based on the reviewed literature, this study develops an AI literacy framework for people of all ages and from all countries. The findings and directions for future research are also discussed.
• Wang, C., & Hu, X. (2025). Exploring the influential factors of online professional learning completion of college teachers from the Global South in an international training program. Internet and Higher Education, 64(Issue). doi:10.1016/j.iheduc.2024.100963
  More info

  Online courses emerged as an important mode for large-scale cross-national teachers' professional learning. However, with most previous research on teacher online professional learning (TOPL) focusing on resource-rich and technology-advanced regions, little attention has been paid to the factors influencing the online learning completion of college teachers in Global South contexts. This study aimed to explore the facilitators and inhibitors of this population's online learning completion in a cross-country program. In seven courses, individual, institutional, and country-level data of 3529 teacher-learners from 99 countries were collected. Forty-two learners were further interviewed. We adopted hierarchical linear modeling to analyze the nested relationships among the individual/institutional/country-level factors and course completion. Results revealed several significant associations between individual/institutional/country-level variables and course completion, as well as several moderation effects. Interviews complemented the analytics results. This study uncovers influential factors of TOPL in Global South contexts and provides practical implications for college teachers' online professional learning.
• Xu, Y., Jiang, T., Hu, X., & Tian, H. (2025). Promoting health behavioral intention through short videos: roles of audiovisual cross-modal correspondence in health communication. Aslib Journal of Information Management. doi:10.1108/ajim-04-2024-0319
  More info

  Purpose: Health short videos are serving as a powerful tool for encouraging individuals to actively adopt healthier behaviors. The sensory cues applied in these videos can be useful for engaging peripheral processing and enhancing attitudes. While previous research has examined the effects of various single cues, this study features a pioneering attempt to explore the roles of audiovisual cross-modal correspondence, encompassing multisensory cues perceived through different modalities, in health communication. Design/methodology/approach: A 2 (color: warm/cool) × 2 (music tempo: fast/slow) between-subjects experiment was conducted to observe 120 participants’ responses to a health short video promoting eye health that was created using four different combinations of background color and background music tempo. Findings: It was found that the congruent color–tempo pairings, that is blue & slow and orange & fast, led to more positive attitudes toward the videos than the incongruent pairings, that is blue & fast and orange & slow. The effect of cross-modal correspondence on attitude was fully mediated by processing fluency, with gender acting as a moderator between the two variables. Furthermore, individuals’ attitudes toward a short video positively influenced their health behavioral intentions. Originality/value: These findings not only lend support to the theoretical framework of “multisensory cues-fluency-attitude-intention” chain for persuasion purposes but also have practical implications for creating effective health short videos.
• Ba, S., Hu, X., Stein, D., & Liu, Q. (2024). Anatomizing online collaborative inquiry using directional epistemic network analysis and trajectory tracking. British Journal of Educational Technology, 55(Issue 5). doi:10.1111/bjet.13441
  More info

  Accurate assessment and effective feedback are crucial for cultivating learners' abilities of collaborative problem-solving and critical thinking in online inquiry-based discussions. Based on quantitative content analysis (QCA), there has been a methodological evolvement from descriptive statistics to sequential mining and to network analysis for mining coded discourse data. Epistemic network analysis (ENA) has recently gained increasing recognition for modelling and visualizing the temporal characteristics of online discussions. However, due to methodological restraints, some valuable information regarding online discussion dynamics remains unexplained, including the directionality of connections between theoretical indicators and the trajectory of thinking development. Guided by the community of inquiry (CoI) model, this study extended generic ENA by incorporating directional connections and stanza-based trajectory tracking. By examining the proposed extensions with discussion data of an online learning course, this study first verified that the extensions are comparable with QCA, indicating acceptable assessment validity. Then, the directional ENA revealed that two-way connections between CoI indicators could vary over time and across groups, reflecting different discussion strategies. Furthermore, trajectory tracking effectively detected and visualized the fine-grained progression of thinking. At the end, we summarize several research and practical implications of the ENA extensions for assessing the learning process.Practitioner notesWhat is already known about this topic Assessment and feedback are crucial for cultivating collaborative problem-solving and critical thinking in online inquiry-based discussions. Cognitive presence is an important construct describing the progression of thinking in online inquiry-based discussions. Epistemic network analysis is gaining increasing recognition for modelling the temporal characteristics of online inquiries. What this paper adds Directional connections between discourses can reflect different online discussion strategies of groups and individuals. A pair of connected discourses coded with the community of inquiry model can have different meanings depending on their temporal order. A trajectory tracking approach can uncover the fine-grained progression of thinking in online inquiry-based discussions. Implications for practice and/or policy Besides the occurrences of individual discourses, examining the meanings of directional co-occurrences of discourses in online discussions is worthwhile. Groups and individuals can employ different discussion strategies and follow diverse paths to thought development. Developmental assessment is crucial for understanding how participants achieve specific outcomes and providing adaptive feedback.
• Liu, Y., Ng, J. T., Hu, X., Ma, Z., & Lai, X. (2024). Adopt or abandon: Facilitators and barriers of in-service teachers’ integration of game learning analytics in K–12 classrooms?. Computers and Education, 209(Issue). doi:10.1016/j.compedu.2023.104951
  More info

  Game learning analytics (GLA) is an emerging technology that facilitates teachers’ evidence-based pedagogical design and assessments. Despite its affordances and potential in K–12 classrooms, teachers’ integration of GLA in teaching practices remains largely unexplored. This study implemented an educational game on collaborative problem solving (CPS) and a GLA system for assisting K–12 teachers in evaluating students’ CPS skills and processes and quest performance and engagement. Based on the integrative model of behavioural prediction, this study aimed to examine 1) the extent to which personal, environmental, and technological factors affected teachers’ usage intention and behaviour towards the GLA system, 2) the effects of moderators on the intention–behaviour relationship, and 3) how the structural model relationships differed across teachers with various individual characteristics. Survey data from 300 in-service teachers from Chinese primary and secondary schools were collected and analysed using partial least squares structural equation modelling. Results indicated that our model demonstrated strong in-sample and out-of-sample predictive power. In particular, teachers’ attitudes, subjective norms, and self-efficacy influenced their behavioural intention, while technological pedagogical content knowledge, school support, and behavioural intention predicted their actual behaviour. In addition, technostress acted as a significant moderator of the intention–behaviour relationship. Moreover, teachers’ gaming preferences, teaching subjects, and years of teaching explained the heterogeneity of their GLA usage. This study contributes to a theoretical understanding of and methodological advancements in studying teachers’ usage intention and behaviour on GLA and yields practical implications for the design and implementation of GLA in K–12 classrooms.
• Park, S. Y., Lee, J. H., Laplante, A., Hu, X., & Kaneshiro, B. (2024). Collaborative Playlists around the World: A Cross-Cultural User Study. Transactions of the International Society for Music Information Retrieval, 7(Issue 1). doi:10.5334/tismir.169
  More info

  Collaborative playlists (CPs) enable users of streaming platforms to share and discover music through co-curation. Recent studies involving predominantly North American samples have found that CPs are created for a variety of contexts, help users organize and access music, facilitate music discovery, and support social connections. Yet, despite these important benefits, little is known about how CP usage aligns or varies across different cultures. We conducted an exploratory study to better understand the landscape of collaborative music engagement with a focus on Hong Kong, South Korea, Quebec, and the United States. We found that across these cultures, previously established purposes for engaging in CPs apply, yet with different degrees of emphasis. Perceived and expected CP outcomes and broader perspectives on social connection through music also varied by location and CP user type. With these findings we discuss primary similarities and differences across the studied cultures and highlight directions for future investigations to further elucidate how music platforms with CP functionalities—and social capabilities more generally—can better help users achieve their desired goals around music.
• Tian, D., Hu, X., Qian, Y., & Li, J. (2024). Exploring the scientific impact of negative results. Journal of Informetrics, 18(Issue 1). doi:10.1016/j.joi.2023.101481
  More info

  Negative results are a routine part of the scientific research journey, yet they often receive insufficient attention in scientific publications. In this study, we investigate the scientific impact of negative results by comparing the citations and citation context between negative and positive results. Specifically, we compared 159 negative result papers from three journals: Journal of Negative Results in BioMedicine, PLoS One, and BMC Research Notes, with 1,058 matched positive result papers authored by the same first and corresponding authors. The citation context was categorized according to three dimensions: citation aspect, citation purpose, and citation polarity. The first two were automatically provided by Citation Opinion Retrieval and Analysis (CORA), while citation polarity was manually annotated. Our analysis revealed several key findings. Firstly, negative results received 38.6 % fewer citations than positive results, even after controlling for bibliographic factors. Secondly, negative results were associated with a significantly higher proportion of negative citations when compared to positive results. Lastly, a higher proportion of negative results were negatively cited in the methods section.
• Ba, S., & Hu, X. (2023). Measuring emotions in education using wearable devices: A systematic review. Computers and Education, 200(Issue). doi:10.1016/j.compedu.2023.104797
  More info

  Wearable devices that detect real-time and fine-grained physiological signals offer potentials for understanding the intricate mechanisms of emotions in education. However, due to the diversities of wearable devices, physiological signals, educational emotions, and educational contexts, there is lack of consensus on the affordance and constraints of wearable devices for measuring emotions in education. The present study conducted a systematic literature review and examined 50 peer-reviewed journal articles and influential proceedings published over the last 15 years (January 2008 to December 2022). Five research questions were addressed concerning research backgrounds, theoretical frameworks, methodologies, remaining challenges, and ethical considerations. Findings demonstrated that while most studies focused on university students in controlled environments, recent advances in wearable devices have enabled emotion measurements of younger learners in natural settings. Research interests have developed towards understanding the theoretical connections between emotion and cognition leveraging wearable devices. Electrodermal activity and heart rate were the most frequently measured signals whereas “engagement”, “positive”, and “anxiety” were the most studied emotions. Machine learning and inferential statistics were often adopted to examine associations between physiological signals and educational emotions. Moreover, we identified a need for updated ethical guidelines in advanced data collection using wearable devices. This review can not only inform wearable device usages in educational practices but also shed light on future research.
• Ba, S., Hu, X., & Law, N. (2023). Daily activities and social interactions predict students’ positive feelings. Asia Pacific Journal of Education, 45(Issue). doi:10.1080/02188791.2023.2219414
  More info

  Positive feelings are essential for students’ well-being and are associated with their academic performance and long-term development. While prior studies have revealed relationships between certain events (e.g., activities and social interactions) and student feelings, little attention was paid to the influence of event durations. In order to address this gap, the present study investigates how time spent on daily activities (e.g., studying) and interactions with social companions (e.g., family/friends) predict adolescent students’ positive feelings. Moreover, the potential moderating roles of personal factors (e.g., health consciousness) were considered. We collected longitudinal data associated with the physical, social, emotional, and digital well-being of 36 middle school students in Hong Kong consecutively for three weeks, using a day reconstruction method. In total, 279 reconstructed days with 2433 events have been recorded. Hierarchical linear modelling was then employed to analyse the nested relationships between events, positive feelings, and personal factors. Results indicated several significant associations between time allocated to daily activities/social interactions and duration of positive feelings. Furthermore, we found that personal factors such as mental health and academic engagement were not only significantly associated with duration of positive feelings but also moderated the relationships between daily activities/social interactions and positive feelings.
• Ba, S., Hu, X., Stein, D., & Liu, Q. (2023). Assessing cognitive presence in online inquiry-based discussion through text classification and epistemic network analysis. British Journal of Educational Technology, 54(Issue 1). doi:10.1111/bjet.13285
  More info

  Providing coaching to participants in inquiry-based online discussions contributes to developing cognitive presence (CP) and higher-order thinking. However, a primary issue limiting quality and timely coaching is instructors' lack of tools to efficiently identify CP phases in massive discussion transcripts and effectively assess learners' cognitive development. This study examined a computational approach integrating text mining and co-occurrence analysis for assessing CP and cognitive development in online discussions based on the community of inquiry (CoI) framework. First, text classifiers trained on different language models were evaluated for identifying and coding the CP phases. Second, epistemic network analysis (ENA) was employed to model cognitive patterns reflected by co-occurrences between the coding elements. Results indicated that text classifiers trained on the state-of-the-art language model Bidirectional Encoder Representations from Transformers (BERT) can address the efficiency issue in coding CP phases in discussion transcripts and obtain substantial agreements (Cohen's k = 0.76) with humans, which outperformed other baseline classifiers. Furthermore, compared to traditional quantitative content analysis, ENA can effectively model the temporal characteristics of online discourse and detect fine-grained cognitive patterns. Overall, the findings suggest a feasible path for applying learning analytics to tracking learning progression and informing theory-based assessments. Practitioner notes What is already known about this topic Cognitive presence is an important construct describing the progression of thinking in online inquiry-based discussions. Most studies used self-report instruments or quantitative content analysis to measure and assess cognitive presence. More efficient and effective approaches were needed by instructors to support assessment of cognitive development and determine coaching strategies. What this paper adds An integrated computational approach for the developmental and formative assessment of cognitive presence was proposed and evaluated. A BERT-based text classification model could efficiently code massive transcripts and achieve substantial agreements with human coders. Epistemic network analysis effectively revealed the process of cognitive development and identified representative discussion patterns and behaviours. Implications for practice and/or policy The proposed approach can considerably reduce the pressure on instructors, enabling them to focus on quality coaching and feedback. Compared to frequencies of individual codes, the connective features between codes carry more insights for assessing cognitive patterns. Learners in a discussion group play different roles and produce diverse paths of cognitive development.
• Haruna, H., Zainuddin, Z., Okoye, K., Mellecker, R. R., Hu, X., Chu, S. K., & Hosseini, S. (2023). Improving instruction and sexual health literacy with serious games and gamification interventions: an outlook to students’ learning outcomes and gender differences. Interactive Learning Environments, 31(Issue 4). doi:10.1080/10494820.2021.1888754
  More info

  This study evaluates the effect of game mechanics by employing serious games and gamification for improving learning outcomes. This is done by considering the performances of the students in terms of the test scores. In the quasi-experimental set-up, which consists of 108 students in their intact three classes attended a series of lessons on sexual health course that was delivered using serious games, gamification, and conventional training elements. The test scores were analysed and compared based on the defined methodology of this paper using SPSS statistical tool version 24. This study found that the students learning outcomes improved after the interventions as the results indicated that there was a statistically significant increase in the test scores from pre-test (M = 24.65, SD = 6.38) to post-test (M = 74.96, SD = 15.89), t(107) = 32.48, p
• Li, F., & Hu, X. (2023). Background Music for Studying: A Naturalistic Experiment on Music Characteristics and User Perception. IEEE Multimedia, 30(Issue 1). doi:10.1109/mmul.2023.3243209
  More info

  Despite the advances in context-aware background music (BM) recommendation, automated BM selection for studying-related contexts is still challenging in that the BM has to not only increase users-activation and task engagement but also avoid distraction. This study investigated how characteristics of BM linked to user's perceptions on task engagement and distraction. In a one-week naturalistic user experiment, 30 participants performed their everyday learning-related tasks with music selected by a BM player. We captured participant's learning contexts and perceptions via pop-up surveys and extracted fine-grained acoustic features for each song in their music listening history via audio processing techniques. Our findings support the power of music in fostering positive studying experience (e.g., perceived engagement) and reveal how several BM characteristics may link to perceived engagement in certain (but not all) conditions. Findings are discussed in relation to theoretical BM studies and implications for generating personalized and context-sensitive BM selections in music-enhanced learning environments.
• Ng, D. T., Lee, M., Tan, R. J., Hu, X., Downie, J. S., & Chu, S. K. (2023). A review of AI teaching and learning from 2000 to 2020. Education and Information Technologies, 28(Issue 7). doi:10.1007/s10639-022-11491-w
  More info

  In recent years, with the popularity of AI technologies in our everyday life, researchers have begun to discuss an emerging term “AI literacy”. However, there is a lack of review to understand how AI teaching and learning (AITL) research looks like over the past two decades to provide the research basis for AI literacy education. To summarize the empirical findings from the literature, this systematic literature review conducts a thematic and content analysis of 49 publications from 2000 to 2020 to pave the way for recent AI literacy education. The related pedagogical models, teaching tools and challenges identified help set the stage for today’s AI literacy. The results show that AITL focused more on computer science education at the university level before 2021. Teaching AI had not become popular in K-12 classrooms at that time due to a lack of age-appropriate teaching tools for scaffolding support. However, the pedagogies learnt from the review are valuable for educators to reflect how they should develop students’ AI literacy today. Educators have adopted collaborative project-based learning approaches, featuring activities like software development, problem-solving, tinkering with robots, and using game elements. However, most of the activities require programming prerequisites and are not ready to scaffold students’ AI understandings. With suitable teaching tools and pedagogical support in recent years, teaching AI shifts from technology-oriented to interdisciplinary design. Moreover, global initiatives have started to include AI literacy in the latest educational standards and strategic initiatives. These findings provide a research foundation to inform educators and researchers the growth of AI literacy education that can help them to design pedagogical strategies and curricula that use suitable technologies to better prepare students to become responsible educated citizens for today’s growing AI economy.
• Qiao, C., & Hu, X. (2023). Leveraging Semantic Facets for Automatic Assessment of Short Free Text Answers. IEEE Transactions on Learning Technologies, 16(Issue 1). doi:10.1109/tlt.2022.3199469
  More info

  Free text answers to short questions can reflect students' mastery of concepts and their relationships relevant to learning objectives. However, automating the assessment of free text answers has been challenging due to the complexity of natural language. Existing studies often predict the scores of free text answers in a 'black box' manner without analyzing their semantic components, which at least partially limit the prediction performance. In this article, we focus on fine-grained semantic facets in free text answers that correspond to knowledge to be mastered. Using a dataset with semantic facet annotation, we first show the correspondence of semantic facet matching states and answer quality, as well as the importance of semantic facets in automatic assessment of answer quality. We then extend the work to a dataset without semantic facet annotation and demonstrate the effectiveness of proposed automated methods in assessing answer quality, including semantic facet extraction, matching state prediction based on a neural framework, and feature engineering with semantic facets. The contribution of this research is twofold: 1) the proposed methods improve state-of-the-art performance of automatic assessment of free text answers and 2) it delves into fine-grained semantic components of free text answers, making it possible to explain the scores and generate detailed feedback.
• Que, Y., Zheng, Y., Hsiao, J. H., & Hu, X. (2023). Studying the effect of self-selected background music on reading task with eye movements. Scientific Reports, 13(Issue 1). doi:10.1038/s41598-023-28426-1
  More info

  Using background music (BGM) during learning is a common behavior, yet whether BGM can facilitate or hinder learning remains inconclusive and the underlying mechanism is largely an open question. This study aims to elucidate the effect of self-selected BGM on reading task for learners with different characteristics. Particularly, learners’ reading task performance, metacognition, and eye movements were examined, in relation to their personal traits including language proficiency, working memory capacity, music experience and personality. Data were collected from a between-subject experiment with 100 non-native English speakers who were randomly assigned into two groups. Those in the experimental group read English passages with music of their own choice played in the background, while those in the control group performed the same task in silence. Results showed no salient differences on passage comprehension accuracy or metacognition between the two groups. Comparisons on fine-grained eye movement measures reveal that BGM imposed heavier cognitive load on post-lexical processes but not on lexical processes. It was also revealed that students with higher English proficiency level or more frequent BGM usage in daily self-learning/reading experienced less cognitive load when reading with their BGM, whereas students with higher working memory capacity (WMC) invested more mental effort than those with lower WMC in the BGM condition. These findings further scientific understanding of how BGM interacts with cognitive tasks in the foreground, and provide practical guidance for learners and learning environment designers on making the most of BGM for instruction and learning.
• Sun, J. C., Liu, Y., Lin, X., & Hu, X. (2023). Temporal learning analytics to explore traces of self-regulated learning behaviors and their associations with learning performance, cognitive load, and student engagement in an asynchronous online course. Frontiers in Psychology, 13(Issue). doi:10.3389/fpsyg.2022.1096337
  More info

  Self-regulated learning (SRL) plays a critical role in asynchronous online courses. In recent years, attention has been focused on identifying student subgroups with different patterns of online SRL behaviors and comparing their learning performance. However, there is limited research leveraging traces of SRL behaviors to detect student subgroups and examine the subgroup differences in cognitive load and student engagement. The current study tracked the engagement of 101 graduate students with SRL-enabling tools integrated into an asynchronous online course. According to the recorded SRL behaviors, this study identified two distinct student subgroups, using sequence analysis and cluster analysis: high SRL (H-SRL) and low SRL (L-SRL) groups. The H-SRL group showed lower extraneous cognitive load and higher learning performance, germane cognitive load, and cognitive engagement than the L-SRL group did. Additionally, this study articulated and compared temporal patterns of online SRL behaviors between the student subgroups combining lag sequential analysis and epistemic network analysis. The results revealed that both groups followed three phases of self-regulation but performed off-task behaviors. Additionally, the H-SRL group preferred activating mastery learning goals to improve ethical knowledge, whereas the L-SRL group preferred choosing performance-avoidance learning goals to pass the unit tests. The H-SRL group invested more in time management and notetaking, whereas the L-SRL group engaged more in surface learning approaches. This study offers researchers both theoretical and methodological insights. Additionally, our research findings help inform practitioners about how to design and deploy personalized SRL interventions in asynchronous online courses.
• Hu, X., Li, F., & Liu, R. (2022). Detecting Music-Induced Emotion Based on Acoustic Analysis and Physiological Sensing: A Multimodal Approach. Applied Sciences (Switzerland), 12(Issue 18). doi:10.3390/app12189354
  More info

  The subjectivity of listeners’ emotional responses to music is at the crux of optimizing emotion-aware music recommendation. To address this challenge, we constructed a new multimodal dataset (“HKU956”) with aligned peripheral physiological signals (i.e., heart rate, skin conductance, blood volume pulse, skin temperature) and self-reported emotion collected from 30 participants, as well as original audio of 956 music pieces listened to by the participants. A comprehensive set of features was extracted from physiological signals using methods in physiological computing. This study then compared performances of three feature sets (i.e., acoustic, physiological, and combined) on the task of classifying music-induced emotion. Moreover, the classifiers were also trained on subgroups of users with different Big-Five personality traits for further customized modeling. The results reveal that (1) physiological features contribute to improving performance on valence classification with statistical significance; (2) classification models built for users in different personality groups could sometimes further improve arousal prediction; and (3) the multimodal classifier outperformed single-modality ones on valence classification for most user groups. This study contributes to designing music retrieval systems which incorporate user physiological data and model listeners’ emotional responses to music in a customized manner.
• Hu, X., Ng, J. T., & Chu, S. K. (2022). Implementing learning analytics in wiki-supported collaborative learning in secondary education: A framework-motivated empirical study. International Journal of Computer-Supported Collaborative Learning, 17(Issue 3). doi:10.1007/s11412-022-09377-7
  More info

  Learning analytics (LA) and group awareness tools are regarded as top priorities for research in the field of computer-supported collaborative learning. As such, this study investigated whether LA-enabled group awareness information facilitates wiki-supported collaborative learning in secondary education. We proposed an analytic framework of measures for assessing collaboration quality in a wiki-based collaborative learning environment, covering student contribution, participation, transactivity, and social dynamics. Based on this framework, we designed an LA-enabled group awareness tool, Wikiglass, for use by both teachers and students in K-12 schools for visualizing statistics of students’ input and interactions on wikis at the class, group, and individual levels. Adopting a naturalistic design, this study allowed teachers and students to decide whether and how often to use the tool. System logs from wikis and Wikiglass and interview data were collected from 440 students and six teachers involved in semester-long wiki-supported group inquiry projects in a secondary school. Regression analyses of quantitative data and thematic content analysis of interview responses showed relationships between the frequencies of teachers’ and students’ use of Wikiglass and measures of students’ collaboration quality at both the individual and group levels. These results indicate that teachers’ scaffolding, students’ collaboration styles, and ethical issues must all be considered when implementing collaborative learning approaches for secondary education. We also discuss the implications of our results for research and practice in the application of LA and group awareness tools for enhancing wiki-supported collaborative learning in K-12 education.
• Li, X., Li, X., Islam, A. Y., Islam, A. Y., Cheng, E. W., Cheng, E. W., Hu, X., Hu, X., Wah Chu, S. K., & Wah Chu, S. K. (2022). Exploring determinants influencing information literacy with activity theory. Online Information Review, 46(Issue 3). doi:10.1108/oir-03-2020-0092
  More info

  Purpose: This study aimed to provide evidence to support the use of a wiki called PBworks as a learning tool to foster students' information literacy (IL) skills based on activity theory. Design/methodology/approach: The participants consisted of 421 students (i.e. form 1 to form 3) from Hong Kong taking a liberal studies course during the 2016–2017 academic year. This study mainly used a mixed methods design, proposing 11 hypotheses. Quantitative data from 374 questionnaires were analysed to test these research hypotheses, while a qualitative method (interviews) was used to explain the quantitative results. A structural equation modelling approach was used to analyse the data, and data triangulation was used to answer the same research questions. Findings: The results showed that the model components PBworks affordances (PB) and rules and divisions (RD) had significant direct effects on individual activities (IA) and community activities (CA) and significant indirect effects on information literacy (IL). The results also revealed that CA had a significant effect on IA and had an even greater effect on IL. Research limitations/implications: Using PBworks and the project-based learning (PjBL) approach, this study examined the determinants affecting the IL skills of Hong Kong junior secondary school students and proposed a wiki-based information literary activity (WILA) model. Practical implications: As students' IL skills have become increasingly important, this study can shed light on related topics for future studies. Social implications: And contribute to social stability and harmonious development. Originality/value: This study eventually confirmed the validity of the WILA model with all hypotheses supported. Peer review: The peer review history for this article is available at: https://publons.com/publon/10.1108/OIR-03-2020-0092.
• Liu, M., & Hu, X. (2022). Movers' advantages: The effect of mobility on scientists' productivity and collaboration. Journal of Informetrics, 16(Issue 3). doi:10.1016/j.joi.2022.101311
  More info

  With the rapid globalization of science, mobility is perceived as an important driver of scientific progress and innovation success. However, we have little knowledge about whether and how scientists' mobility influences their career development, especially scientists' productivity and collaboration. In this case study, using the data on 62,330 scientists, the Chinese computer scientists who published at least one computer science paper and published no fewer than 10 papers in total from 2000 to 2012, we apply difference in differences models in conjunction with PSM methods to show the effect of domestic mobility (i.e., moving inside China) on scientists' research quantity and quality by distinguishing the direction of mobility. In contrast to the existing literature that documents a short-term negative effect due to adaption costs or disruption of routines and social capital, we do not observe an initial detrimental impact of following moves on productivity and collaboration, even for non-upward moves. We further find that mobility leads to increased collaboration with new partners without dampening scientists' collaboration with previous collaborators. However, scientists have a higher probability of collaborating with new collaborators, as evidenced by the decreased share of previous collaborators to the total co-authors after they move. The findings of this case study imply that the benefits of mobility might outweigh its costs and that mobility improves scientists' productivity and collaboration for prolific scientists in emerging countries.
• Liu, M., Zhang, N., Hu, X., Jaiswal, A., Xu, J., Chen, H., Ding, Y., & Bu, Y. (2022). Further divided gender gaps in research productivity and collaboration during the COVID-19 pandemic: Evidence from coronavirus-related literature. Journal of Informetrics, 16(Issue 2). doi:10.1016/j.joi.2022.101295
  More info

  Based on publication data on coronavirus-related fields, this study applies a difference in differences approach to explore the evolution of gender inequalities before and during the COVID-19 pandemic by comparing the differences in the numbers and shares of authorships, leadership in publications, gender composition of collaboration, and scientific impacts. We find that, during the pandemic: (1) females' leadership in publications as the first author was negatively affected; (2) although both females and males published more papers relative to the pre-pandemic period, the gender gaps in the share of authorships have been strengthened due to the larger increase in males' authorships; (3) the share of publications by mixed-gender collaboration declined; (4) papers by teams in which females play a key role were less cited in the pre-pandemic period, and this citation disadvantage was exacerbated during the pandemic; and (5) gender inequalities regarding authorships and collaboration were enhanced in the initial stage of COVID-19, widened with the increasing severity of COVID-19, and returned to the pre-pandemic level in September 2020. This study shows that females' lower participation in teams as major contributors and less collaboration with their male colleagues also reflect their underrepresentation in science in the pandemic period. This investigation significantly deepens our understanding of how the pandemic influenced academia, based on which science policies and gender policy changes are proposed to mitigate the gender gaps.
• Gomez-Canon, J. S., Cano, E., Eerola, T., Herrera, P., Hu, X., Yang, Y. H., & Gomez, E. (2021). Music Emotion Recognition: Toward new, robust standards in personalized and context-sensitive applications. IEEE Signal Processing Magazine, 38(Issue 6). doi:10.1109/msp.2021.3106232
  More info

  Emotion is one of the main reasons why people engage and interact with music [1]. Songs can express our inner feelings, produce goosebumps, bring us to tears, share an emotional state with a composer or performer, or trigger specific memories. Interest in a deeper understanding of the relationship between music and emotion has motivated researchers from various areas of knowledge for decades [2], including computational researchers. Imagine an algorithm capable of predicting the emotions that a listener perceives in a musical piece, or one that dynamically generates music that adapts to the mood of a conversation in a film - a particularly fascinating and provocative idea. These algorithms typify music emotion recognition (MER), a computational task that attempts to automatically recognize either the emotional content in music or the emotions induced by music to the listener [3]. To do so, emotionally relevant features are extracted from music. The features are processed, evaluated, and then associated with certain emotions. MER is one of the most challenging high-level music description problems in music information retrieval (MIR), an interdisciplinary research field that focuses on the development of computational systems to help humans better understand music collections. MIR integrates concepts and methodologies from several disciplines, including music theory, music psychology, neuroscience, signal processing, and machine learning.
• Haruna, H., Okoye, K., Zainuddin, Z., Hu, X., Chu, S., & Hosseini, S. (2021). Gamifying sexual education for adolescents in a low-Tech setting: Quasi-experimental design study. JMIR Serious Games, 9(Issue 4). doi:10.2196/19614
  More info

  Background: Sexual education has become increasingly important as unhealthy sexual practices and subsequent health risksbecome more prevalent during adolescence. Traditional sex education teaching methodologies are limiting for digital nativesexposed to various digital technologies. Harnessing the power of technology applications attractive to the younger generationmay be a useful approach for teaching sex education.Objective: The aim of this study was to improve sexual health knowledge and understanding of the problems associated withunhealthy sexual practices and address sexual and reproductive health challenges experienced in a low-Tech setting.Methods: A participatory design approach was used to develop the digital gamified methodology. A sample of 120 secondaryschool students aged 11-15 were randomly assigned to either experimental or control group for each of the 3 teaching approaches:(1) gamified instruction (actual serious games [SG] in teaching); (2) gamification (GM; making nongames, such as game-likelearning); and (3) traditional teaching (TT) methods.Results: The SG and GM approaches were more effective than TT methods in teaching sexual health education. Specifically,the average scores across groups demonstrated an increase of mean scores from the pre-to posttest (25.10 [SD 5.50] versus 75.86[SD 13.16]; t119=41.252; P
• Hu, X., Hu, X., Chen, J., Chen, J., Wang, Y., & Wang, Y. (2021). University students’ use of music for learning and well-being: A qualitative study and design implications. Information Processing and Management, 58(Issue 1). doi:10.1016/j.ipm.2020.102409
  More info

  Music has long been recognised to be able to alter people's emotions and behaviours, yet how university students use music for learning and well-being is largely unexplored. With one of the largest music user populations in the world, China has tremendous market potential for digital music. This study explores music use behaviours for learning and well-being among university students in China and how these findings can inform future online music service design. An investigation framework is developed based on theories in multiple related disciplines such as musicology, psychology, and sociology. In-depth interviews were conducted with forty university students in twenty universities, with an interview protocol designed based on the framework. Interview transcriptions were analysed using a thematic content analysis approach. The results reveal how students use music for multiple aspects of life corresponding to learning and major components of well-being, including physical well-being, social relationships, positive emotion, self-esteem, and meaning of life. Based on the findings we discuss emerging themes on the design of online music information systems and services. This study fills the research gap on how music is used by university students for benefiting learning and well-being. The design implications are valuable for online music services to better meet users’ evolving needs. The proposed framework and method can be readily used to study music users in various populations.
• Lin, Y. T., Liao, Y. Z., Hu, X., & Wu, C. C. (2021). EEG Activities during Program Comprehension: An Exploration of Cognition. IEEE Access, 9(Issue). doi:10.1109/access.2021.3107795
  More info

  This study attempts to explore cognition during program comprehension through physiological evidence by recording and comparing electroencephalogram (EEG) activities in different frequency bands and the eye movements of the participants with high or low programming abilities. An experiment was conducted with thirty-three undergraduate students majoring in Computer Science. We recorded their EEG activities when they were reading two programs with three types of program constructs. At the same time, the participants' eye movements were recorded by an eye tracker to further understand the relationship between the program comprehension process and EEG activities. Experimental results show that the high-performance participants displayed higher performance for working memory (theta power), attention resource allocation (lower alpha power), and interaction between working memory and semantic memory (upper alpha power) in program comprehension tasks of complex constructs, which proves related theories proposed in the existing research on programming and cognition. The results of this study not only offer objective evidence of the roles cognition plays in program comprehension but also provide educators with suggestions for designing suitable pedagogical strategies.
• Liu, M., & Hu, X. (2021). Will collaborators make scientists move? A Generalized Propensity Score analysis. Journal of Informetrics, 15(Issue 1). doi:10.1016/j.joi.2020.101113
  More info

  Through collaboration, scientists' human and social capital are accumulated that are considered important in the academic job market. However, little is known about whether academic past collaboration influence scientists' mobility. To deal with treatment endogeneity, we conduct a Generalized Propensity Score analysis (GPS) and apply a novel application of the Dose-Response Function model. Using the data on 15,968 Chinese scientists from 2000 to 2012 as an illustrative case, we find that 1) the number of domestic and overseas collaborators are positively associated with scientists' mobility and upward move, while the magnitude of the effect of overseas collaborators is far smaller than that of domestic collaborators; 2) domestic collaborators' productivity is positively related to scientists' move and upward move; 3) there is a stronger effect of collaborators from higher-tier universities on scientists' upward move; 4) we do not observe a significant relationship between the recent stock of collaborators and scientists' mobility. In addition to implications for talent policies and scientists' career development, this study makes significant methodological contributions through introducing a new method, GPS, to address selection bias of the independent variable, i.e., scientists' collaboration. Our results show that, with great potential to capture causality, GPS facilitates research in informetrics, scientometrics and science policy from a quantitative perspective, and enriches policy relevance of the findings.
• Zou, W., Hu, X., Pan, Z., Li, C., Cai, Y., & Liu, M. (2021). Exploring the relationship between social presence and learners’ prestige in MOOC discussion forums using automated content analysis and social network analysis. Computers in Human Behavior, 115(Issue). doi:10.1016/j.chb.2020.106582
  More info

  Research has repeatedly proven the importance of social interactions in online learning contexts such as Massive Open Online Courses (MOOCs), where learners often reported isolation and a lack of peer support. Previous studies of social presence suggested that the ways learners present themselves socially online affect their learning outcomes. In order to further understand the role of learners' social presence, this study attempts to examine the relationship between social presence and learners' prestige in the learner network of a MOOC. An automated text classification model based on the latest machine learning techniques was developed to identify different social presence indicators from forum posts, while two metrics (in-degree and authority score) in social network analysis (SNA) were used to measure learners' prestige in the learner network. Results revealed that certain social presence indicators such as Asking questions, Expressing gratitude, Self-disclosure, Sharing resources and Using Vocatives have positive correlations with learners' prestige, while the expressions of Disagreement/doubts/criticism and Negative emotions were counterproductive to learners' prestige. The findings not only reinforce the importance of social presence in online learning, but also shed light on the strategies of leveraging social presence to improve individual's prestige in social learning contexts like MOOCs.
• Chu, S. K., Hu, X., & Ng, J. (2020). Exploring secondary school students’ self-perception and actual understanding of plagiarism. Journal of Librarianship and Information Science, 52(Issue 3). doi:10.1177/0961000619872527
  More info

  Plagiarism has been a growing concern among institutions and academics in recent years. To address the problem, and to alleviate the growing trend of this academic misconduct, students’ perceptions of plagiarism should be considered. This study explores students’ self-perception and actual understanding of plagiarism, and the relations between them. Survey responses were collected from 433 students in a Hong Kong junior secondary school. Results reveal that students show different understanding towards ‘obvious’ and ‘obscure’ plagiarism, with misunderstanding or misconception more likely arising over obscure plagiarism. This study also reports that students’ self-perception on their understanding of plagiarism differed across grade levels, and their academic performance of inquiry-based learning has a relation to their self-perceived and actual understanding of plagiarism. Implications for improving the teaching and learning of plagiarism are discussed.
• Haruna, H., Abbas, A., Zainuddin, Z., Hu, X., Mellecker, R. R., & Hosseini, S. (2020). Enhancing instructional outcomes with a serious gamified system: a qualitative investigation of student perceptions. Information and Learning Science, 12(Issue 5-6). doi:10.1108/ils-05-2020-0162
  More info

  Purpose: This paper aims to evaluate the students’ perception of their learning experiences concerning serious gaming and gamification instructions and determines whether they were motivated enough and engaged during the educative process in a resource-poor context. Moreover, the study evaluated the impact of interactive instructional environment outcomes in terms of students’ perceptions of the learning catalysed by gamified systems, particularly in enhancing attitude change coupled with knowledge acquisition. Design/methodology/approach: This study used a qualitative research design technique to collect the data. A total of 108 first year secondary school students participated in a sexual health literacy course that lasted for a five-week learning period. Using a cluster-sampling technique, three classes were randomly assigned to serious gaming, gamification and teacher-centred instructions. Individual face-to-face interviews were used to assess students’ perceives required satisfaction with three instructions. Data were audio-recorded, and coding analysis was used using NVivo software facilitated qualitative data analysis. Findings: The results show that serious gaming and gamification instructions trumped the traditional teacher-centred instruction method. While intervention students were all positive about the serious gaming and gamification instructions, non-intervention students were negative about conservative teacher-centered learning whose limited interactivity also undermined learning relative to the two innovative interventions. Research limitations/implications: As a justification to limit face-to-face classes, this study may be useful during an emergency phenomenon, including the current situation of amid COVID-19. The implementation of serious gaming and gamification as remotely instructional options could be among the measures to protect educational communities through reducing close-proximity, and eventually, control contamination and the spread of viruses. Originality/value: The application of serious gaming and game elements should not be conceptualised as universal but context-specific. This study shows that particularism is essential to optimise the results in terms of coming up with a specific design based on the scope of evaluation for positive results and develop an intervention that will work, especially in the resource-poor context of the developing world.
• Hu, X., Ng, J., Tsang, K. K., & Chu, S. K. (2020). Integrating Mobile Learning to Learning Management System in Community College. Community College Journal of Research and Practice, 44(Issue 10-12). doi:10.1080/10668926.2019.1640146
  More info

  This article reports findings from a study that aims to understand how to integrate mobile-learning into Learning Management System (LMS) in a community college in Hong Kong. In this study, a mobile-enabled LMS named SOUL was adopted to improve students’ learning engagement and academic performance. Participating students were segregated into two groups where one was prompted by the instructor to use mobile access while another group was not prompted. A survey was conducted to investigate students’ use of SOUL via mobile access and the factors influencing their adoption of mobile access to SOUL based on the Unified Theory of Acceptance and Use of Technology (UTAUT) model. Semi-structured interviews were conducted with students to collect in-depth explanations regarding their experience with SOUL via mobile access. The study reveals that many students used SOUL with their mobile devices despite not being prompted. It was also found that students most frequently accessed SOUL via their mobile devices for retrieving learning resources and information owing to immediate availability and convenience brought by the notification of its native mobile app. Multiple linear regression analyses revealed that facilitating conditions and performance expectancy were the only significant predictor for unprompted and prompted mobile access, respectively. Implications on integrating mobile learning to LMS in community colleges are discussed.
• Liu, M., Zangerle, E., Hu, X., Melchiorre, A., & Schedl, M. (2020). PANDEMICS, MUSIC, AND COLLECTIVE SENTIMENT: EVIDENCE FROM THE OUTBREAK OF COVID-19. Proceedings of the International Society for Music Information Retrieval Conference.
  More info

  The COVID-19 pandemic causes a massive global health crisis and produces substantial economic and social distress, which in turn may cause stress and anxiety among people. Real-world events play a key role in shaping collective sentiment in a society. As people listen to music daily everywhere in the world, the sentiment of music being listened to can reflect the mood of the listeners and serve as a measure of collective sentiment. However, the exact relationship between real-world events and the sentiment of music being listened to is not clear. Driven by this research gap, we use the unexpected outbreak of COVID-19 as a natural experiment to explore how users’ sentiment of music being listened to evolves before and during the outbreak of the pandemic. We employ causal inference approaches on an extended version of the LFM-1b dataset of listening events shared on Last.fm, to examine the impact of the pandemic on the sentiment of music listened to by users in different countries. We find that, after the first COVID-19 case in a country was confirmed, the sentiment of artists users listened to becomes more negative. This negative effect is pronounced for males while females’ music emotion is less influenced by the outbreak of the COVID-19 pandemic. We further find a negative association between the number of new weekly COVID-19 cases and users’ music sentiment. Our results provide empirical evidence that public sentiment can be monitored based on collective music listening behaviors, which can contribute to research in related disciplines.
• Ng, J., Lei, L., Iseli-Chan, N., Li, J., Siu, F., Chu, S., & Hu, X. (2020). Non-repository uses of learning management system through mobile access. Journal of Educational Technology Development and Exchange, 13(Issue 1). doi:10.18785/jetde.1301.01
  More info

  Learning Management Systems (LMSs) have been widely adopted in higher education worldwide, but predominately used as repositories of learning materials. Mobile access to LMSs enables greater mobility and flexible learning, and thus may help boosting non-repository uses of LMSs, maximizing their educational affordance. This study examined the extent to which mobile access to an LMS, Moodle, was used for various learning activities, with a focus on those beyond storing and retrieving learning materials, as well as the factors influencing students’ non-repository uses of LMS via mobile access. A mixed-method approach was applied, with survey responses collected from 316 students and interviews with 26 students and five instructors across nine courses in a comprehensive university in Hong Kong. The results showed that mobile access to non-repository uses of Moodle was significantly less frequent than that to repository uses across all courses, and students viewed mobile access to the Moodle platform largely as a backup to supplement computer access. Findings suggested four inter-related factors influencing mobile access to LMS for non-repository uses, including course LMS activity design, instructors’ attitudes towards LMS, the nature of tasks conducted with LMS, and situational contexts.
• Qiao, C., & Hu, X. (2020). A joint neural network model for combining heterogeneous user data sources: An example of at-risk student prediction. Journal of the Association for Information Science and Technology, 71(Issue 10). doi:10.1002/asi.24322
  More info

  Information service providers often require evidence from multiple, heterogeneous information sources to better characterize users and offer personalized service. In many cases, statistic information (for example, users' profiles) and sequentially dynamic information (for example, logs of interaction with information systems) are two prominent sources that can be combined to achieve optimized results. Previous attempts in combining these two sources mainly exploited models designed for either static or sequential information, but not both. This study aims to fill the gap by proposing a novel joint neural network model that can naturally fit both static and sequential user data. To evaluate the effectiveness of the proposed method, this study uses the problem of at-risk student prediction as an example where both static data (personal profiles) and sequential data (event logs) are involved. A thorough evaluation was conducted on an open data set, with comparisons to a range of existing approaches including both static and sequential models. The results reveal superb performances of the proposed method. Implications of the findings on further research and applications of joint models are discussed.
• Qiao, C., Qiao, C., Hu, X., & Hu, X. (2020). A neural knowledge graph evaluator: Combining structural and semantic evidence of knowledge graphs for predicting supportive knowledge in scientific QA. Information Processing and Management, 57(Issue 6). doi:10.1016/j.ipm.2020.102309
  More info

  Effectively detecting supportive knowledge of answers is a fundamental step towards automated question answering. While pre-trained semantic vectors for texts have enabled semantic computation for background-answer pairs, they are limited in representing structured knowledge relevant for question answering. Recent studies have shown interests in enrolling structured knowledge graphs for text processing, however, their focus was more on semantics than on graph structure. This study, by contrast, takes a special interest in exploring the structural patterns of knowledge graphs. Inspired by human cognitive processes, we propose novel methods of feature extraction for capturing the local and global structural information of knowledge graphs. These features not only exhibit good indicative power, but can also facilitate text analysis with explainable meanings. Moreover, aiming to better combine structural and semantic evidence for prediction, we propose a Neural Knowledge Graph Evaluator (NKGE) which showed superior performance over existing methods. Our contributions include a novel set of interpretable structural features and the effective NKGE for compatibility evaluation between knowledge graphs. The methods of feature extraction and the structural patterns indicated by the features may also provide insights for related studies in computational modeling and processing of knowledge.
• Tavernier, M., & Hu, X. (2020). Emerging Mobile Learning Pedagogy Practices: Using tablets and constructive apps in early childhood education. Educational Media International, 57(Issue 3). doi:10.1080/09523987.2020.1824423
  More info

  Early childhood teachers introduced mobile learning activities and educational software to the children's in-class learning activities. This qualitative study implemented one constructive creation app in two early childhood classrooms. It designed, implemented and evaluated the effectiveness of pedagogy practices (PP) for the meaningful implementation of constructive apps in an early childhood classroom. Two small groups of four to five years-old children engaged in weekly in-class research activities. Sixteen videos and 122 artifacts were analysed to determine each child's well-being, involvement, and motivation during the digital creation activities and evaluate the effectiveness of the implemented PP. The most effective PP embedded strategies that 1. enhanced the children's perceived autonomy, competence, and relatedness, 2. implemented a flexible and generous use of direct interactions between teachers and students, and 3. provided app operational habit-shaping routines that guided young children’s engagement with the mobile device and app. The implications of these findings are practical and theoretical. Teachers may use the transferable PP to guide their efforts to develop and implement digital creation activities in their early childhood classrooms. The findings of this study may also contribute the development of a mobile learning theory for young children and address a gap in the research literature.
• Haruna, H., Hu, X., Chu, S. K., & Mellecker, R. R. (2019). Initial validation of the MAKE framework: A comprehensive instrument for evaluating the efficacy of game-based learning and gamification in adolescent sexual health literacy. Annals of Global Health, 85(Issue 1). doi:10.5334/aogh.1110
  More info

  Objectives: When evaluating the effectiveness of a method for instructing adolescents in sexual health literacy, it is essential to consider how the method motivates learning, promotes a change of attitude, increases knowledge gain, and engages students (MAKE). This article reports on the development and validation of a unified, comprehensive framework for evaluating the efficacy of games in teaching sexual health behaviors for curbing unhealthy sexual outcomes to secondary school adolescents in low resource settings. Methods: The initial validation of the MAKE framework was administered to 120 students using quantitative data collection and analysis. It was then subjected to factor analysis tests to investigate the items’ structure, and Cronbach’s alpha was applied to measure the scale reliability using SPSS Version 24. Results: Data analyses demonstrate that the MAKE framework is a comprehensive instrument to evaluate teaching methods with four powerful constructs, each of which has two to four components. For each construct, the following data were obtained: for motivation, standardized alpha = 0.92, Kaiser-Meyer- Olkin (KMO) = 0.88, and p = 0.001; for attitude, standardized Cronbach’s alpha = 0.90, KMO = 0.88, and p = 0.001; for knowledge, standardized alpha = 0.92, KMO = 0.86, and p = 0.001; and finally, for engagement, standardized alpha = 0.90, KMO = 0.87, and p = 0.001. Cronbach’s alpha for each component was above the cut-off point (0.65). Conclusions: This study shows that the MAKE framework is a satisfactory instrument for assessing the efficacy of teaching methods for sexual health literacy in a variety of teaching environments. The method may also have value for assessing the effectiveness of other methods in adolescent sexual health education.
• Haruna, H., Zainuddin, Z., Mellecker, R. R., Chu, S. K., & Hu, X. (2019). An iterative process for developing digital gamified sexual health education for adolescent students in low-tech settings. Information and Learning Science, 120(Issue 11-12). doi:10.1108/ils-07-2019-0066
  More info

  Purpose: Digital technology has great potential for educating today’s digitally oriented adolescents on health. In particular, digital health gamified learning can make the promotion of the sexual well-being of adolescents more effective. Although venereal diseases such as HIV/AIDS have become a greater problem in Sub-Saharan African (SSA) countries than in any country outside of Africa, little is publicly known about the development of gamified learning for use in counter-measures. This paper aims to address that deficit by presenting the process of developing one such game. The paper highlights how the “My Future Begins Today” game for sexual health education was developed, evaluated and refined in the real-world of low-tech settings and made improvements based on the response of users. Design/methodology/approach: Design-based research (DBR) was used to guide the design, develop, test and refine the digital game in iterative cycles. The evaluation of the effectiveness of iterations of the game was done using adolescent sexual health literacy tests and the validated Motivation, Attitude, Knowledge and Engagement framework, the authors developed based on existing approaches. That framework combines the elements of motivation, attitude, knowledge and engagement, effectiveness was evaluated based on the game’s ability to motivate students, improve their attitudes, increase their acquisition of knowledge and engage them in learning self-rating surveys and interviews. The whole process of game design, testing, evaluation and refinement were underpinned by the activity theory, DBR and participatory design (PD) research. Findings: Participants in the gamified learning platforms demonstrated higher average scores on their post-tests than their counterparts subjected to the traditional teaching classroom. Also, gamified learning groups commented positively on the effectiveness of their instructional approach than their counterparts in the traditional learning group. The stakeholders’ involvement in developing gamified learning provided a good understanding of the importance of the game to the adolescent students and how it was going to be used to address the problem identified. The application of PD contributed to the effectiveness of the game. It involved various actors from various fields who were relevant to the game. Also, engaging targeted users from the beginning resulted in the creation of a better correspondence with the preferences of end-users. Practical implications: This study has contributed to a better understanding of sex education and knowledge in the area of adolescent reproductive health issues, using developed innovative game mechanics features and its applicability in low-tech settings. Originality/value: The study will be a recommendation for future researchers in applying this gamified learning concept and its suitability in their teaching practice, particularly regarding sexual health education and adolescent reproductive health issues in low-tech settings of SSA.
• Hu, X. (2019). Evaluating mobile music services in China: An exploration in user experience. Journal of Information Science, 45(Issue 1). doi:10.1177/0165551518762070
  More info

  Most digital music repositories and services have mobile applications (apps) that facilitate convenient access for users via smartphones. Although China has one of the largest music listener populations in the world, there is little research evaluating Chinese online or mobile music services. To bridge this gap, this study evaluated mobile apps of three of the most popular Chinese music services from the user’s perspective, using usability testing and semi-structured interviews with a sample of active users in China. Nielsen’s 10 user experience heuristics and four criteria in recommender evaluation were examined. Results identified criteria that create a positive user experience, and those that need further improvement. This study contributes to the literature in user-centred evaluation in music information retrieval (MIR) and music digital libraries (MDL), and provides practical insights for music application design, use and evaluation.
• Hu, X., & Lai, C. (2019). Comparing factors that influence learning management systems use on computers and on mobile. Information and Learning Science, 120(Issue 7-8). doi:10.1108/ils-12-2018-0127
  More info

  Purpose: Learning management systems (LMSs) have been embraced for their potential to create a ubiquitous learning that is free from time and space constraints. Mobile devices afford enhanced mobility that enables flexible learning with LMSs. Thus, understanding students’ use of mobile devices to interact with LMSs and the influencing factors is essential. This paper aims to examine the factors that influenced students’ behavioural intention in using Web-based LMSs via mobile phones and compared the factors with those that affect students’ general acceptance of Web-based LMSs. Design/methodology/approach: This study surveyed 356 university students and interviewed 17 students on the various factors that might affect their LMS adoption. Structural equation modelling was used to analyse the survey data. Findings: This study identified that perceived usefulness, perceived ease of use, social influence and facilitating conditions were significant determinants of students’ usage intention in both contexts. However, social factors exerted greater influence on students’ behavioural intentions of mobile access than the attitudinal factors. The results also pinpointed some sociocultural and tempo-spatial factors that might have minimized the influence of perceived usefulness in the mobile context. Originality/value: The study calls for special attention to the potential influences of sociocultural norms and tempo-spatial circumstances of mobile use in shaping the nature of learners’ voluntary mobile use of LMSs.
• Hu, X., Hennebry, M. L., & Cheung, C. L. (2019). Language learning motivation of students from a special educational school in Hong Kong. Asian Journal of Applied Linguistics, 6(Issue 1).
  More info

  Special educational needs (SEN) have attracted considerable attention in education and educational research. Nevertheless, limited research attention has been given to the second language (L2) learning of students with SEN in special schools and even less to their L2 learning motivation (LLM), despite the significant role of LLM in L2 learning success. This paper compares LLM data gathered from 66, grade 7-10, students with SEN in a special school and 66, grade 7-10, non-SEN students from mainstream schools in Hong Kong. Findings from a motivational questionnaire reveal higher levels of LLM among the special school students (SSS) than the mainstream school students (MSS). One-way MANOVA and Cohen's d, show students from the special school have significantly higher ought-to L2 self and English learning attitude yet significantly lower required orientation than their mainstream school peers. Regression analysis has allowed the investigation of factors interacting with LLM among SSS and MSS, suggesting that self-efficacy and parental influence are significant predictors of the LLM for both groups.
• Hu, X., Ng, J., & Lee, J. H. (2019). VR creation experience in cultural heritage education: A preliminary exploration. Proceedings of the Association for Information Science and Technology, 56(Issue 1). doi:10.1002/pra2.42
  More info

  Despite an increasing adoption of virtual reality (VR) in education, few studies have explored VR creation in cultural heritage education. This study aims to investigate students' experience of creating VR content featuring cultural heritage in an undergraduate general education course repurposed from a digital collection course. A survey was conducted with 87 students, collecting both close-ended and open-ended responses. A coding framework was designed to analyze students' open-ended responses. Preliminary findings show that students were largely satisfied with the experience, which helped learners from diverse academic backgrounds toward acquiring technological skills that are essential for the 21st century workforce. The VR creation experience also motivated the students to learn more about cultural heritage. Technical issues regarding spherical photo-taking and the online VR creation tool were identified, which may call for new alternative tools and devices. Findings offer empirical evidence on the value of integrating VR creation into cultural heritage education, as well as implications for pedagogical design and educational applications of VR creation.
• Hu, X., Shan, J., Chen, J., & Downie, J. S. (2019). Affording cross-cultural access to visualized cultural collections: A preliminary case study of e-dunhuang. Proceedings of the Association for Information Science and Technology, 56(Issue 1). doi:10.1002/pra2.128
  More info

  E-Dunhuang is a visual material centered digital library of a UNESCO world heritage. This study aims to identify how users in different cultural backgrounds seek and use cultural heritage information on e-Dunhuang, as well as their perceptions and opinions toward that. As a preliminary report, this poster focuses on the visual content (i.e. images, panorama) of e-Dunhuang, which is arguably the most prominent component in this platform. Results of usability tests and follow-up interviews reveal that users had polarized opinions. The findings can inform cross-cultural access to cultural heritage collections dominant with visual content.
• Ng, J., Hu, X., Luo, M., & Chu, S. K. (2019). Relations among participation, fairness and performance in collaborative learning with Wiki-based analytics. Proceedings of the Association for Information Science and Technology, 56(Issue 1). doi:10.1002/pra2.48
  More info

  Utilizing data analytics for supporting collaborative learning is under-studied in secondary education. This study aims to evaluate the effectiveness of Wiki and Wiki-based learning analytics in facilitating collaborative learning in a junior secondary school, in terms of students' participation, contribution, performance, and perception. A Wiki-based learning analytic tool, Wikiglass, was employed for visualizing statistics of students' contributions and participation (e.g., revision counts) in Wiki, on both the group and individual levels. System log, student survey and performance data were collected from students involved in a Wiki-supported inquiry project assignment. An Unfairness Index is proposed to measure students' group work distribution. Results of statistical analyses show that fairness of group work distribution was positively related to active participation in revisions on Wiki on the group level, and the number of sentences with higher-order thinking was related to group performance scores. On the level of individual students, Wiki-based analytics increased the visibility of work distribution and peers' work progress and contributions which might have changed students' collaborative behaviors.
• Qiao, C., & Hu, X. (2019). Text classification for cognitive domains: A case using lexical, syntactic and semantic features. Journal of Information Science, 45(Issue 4). doi:10.1177/0165551518802522
  More info

  Various automated classifiers have been implemented to categorise learning-related texts into cognitive domains. However, existing studies have applied limited linguistic features, and most have focused on texts written in English, with little attention given to Chinese. This study has tried to fill the gaps by applying a comprehensive set of features that have rarely been used collectively in previous research, with a focus on Chinese analytical texts. Experiments were conducted for classifier learning and evaluation, where a feature selection procedure significantly improved the classification performance. The results showed that different types of features complemented each other in forming strong collective representations of the original texts, and the discriminant nature of the features can be reasonably explained by language usage phenomena. The proposed approach could potentially be applied to other datasets of analytical writings involving cognitive domains, and the text features explored could be reused and further refined in future studies.
• Yu, B., & Hu, X. (2019). Toward training and assessing reproducible data analysis in data science education. Data Intelligence, 1(Issue 4). doi:10.1162/dint_a_00053
  More info

  Reproducibility is a cornerstone of scientific research. Data science is not an exception. In recent years scientists were concerned about a large number of irreproducible studies. Such reproducibility crisis in science could severely undermine public trust in science and science-based public policy. Recent efforts to promote reproducible research mainly focused on matured scientists and much less on student training. In this study, we conducted action research on students in data science to evaluate to what extent students are ready for communicating reproducible data analysis. The results show that although two-thirds of the students claimed they were able to reproduce results in peer reports, only one-third of reports provided all necessary information for replication. The actual replication results also include conflicting claims; some lacked comparisons of original and replication results, indicating that some students did not share a consistent understanding of what reproducibility means and how to report replication results. The findings suggest that more training is needed to help data science students communicating reproducible data analysis.
• Haruna, H., & Hu, X. (2018). International Trends in Designing Electronic Health Information Literacy for Health Sciences Students: A Systematic Review of the Literature. Journal of Academic Librarianship, 44(Issue 2). doi:10.1016/j.acalib.2017.12.004
  More info

  The Internet has become a crucial source of health information for health sciences students. They increasingly rely on the Internet for health information to support their educational projects, academic activities, clinical practice and research. Surprisingly, it has been shown that students' health information skills for conducting research on the Internet are inadequate. Indeed, developing and improving the health information skill set of health sciences students is required in order for students to effectively locate, critically evaluate, and efficiently use online health information for the effective location, critical evaluation and efficient use of online health information. This paper undertakes a systematic review of the literature with a focus on electronic health information literacy skills with the aim of identifying the current trends, contributions to, and practices in health sciences students' education, and informing researchers in the field universally about the essential baseline for the design and development of effective course contents, pedagogy and assessment approaches. However, majority of students have limited skills for the location, evaluation and effective use of health information on the Internet. Other articles suggest that health sciences students need fully fledged health information skills programs that are integrated with their health sciences education curricula.
• Haruna, H., Hu, X., Chu, S. K., Mellecker, R. R., Gabriel, G., & Ndekao, P. S. (2018). Improving sexual health education programs for adolescent students through game-based learning and gamification. International Journal of Environmental Research and Public Health, 15(Issue 9). doi:10.3390/ijerph15092027
  More info

  An effective innovative pedagogy for sexual health education is required to meet the demands of technology savvy digital natives. This study investigates the extent to which game-based learning (GBL) and gamification could improve the sexual health education of adolescent students. We conducted a randomized control trial of GBL and gamification experimental conditions. We made a comparison with traditional teaching as a control condition in order to establish differences between the three teaching conditions. The sexual health education topics were delivered in a masked fashion, 40-min a week for five weeks. A mixed-method research approach was uses to assess and analyze the results for 120 students from a secondary school in Dar Es Salaam, Tanzania. Students were divided into groups of 40 for each of the three teaching methods: GBL, gamification, and the control group (the traditional teaching method). The average post-test scores for GBL (Mean = 79.94, SD = 11.169) and gamification (Mean = 79.23, SD = 9.186) were significantly higher than the control group Mean = 51.93, SD = 18.705 (F (2, 117) = 54.75, p = 0.001). Overall, statistically significant differences (p ≤ 0.05) were found for the constructs of Motivation, Attitude, Knowledge, and Engagement (MAKE). This study suggests that the two innovative teaching approaches can be used to improve the sexual health education of adolescent students. The methods can potentially contribute socially, particularly in improving sexual health behaviour and adolescents’ knowledge in regions plagued by years of sexual health problems, including HIV/AIDS.
• Hu, X. (2018). Usability Evaluation of E-Dunhuang Cultural Heritage Digital Library. Data and Information Management, 2(Issue 2). doi:10.2478/dim-2018-0008
  More info

  Digital libraries have been strategic in preserving and making non-movable cultural heritage information accessible to everyone with network connections. In light of their cultural and historical importance in the ancient “Silk Road,” murals and stone caves in Dunhuang, a remote city in northwest China,have been digitized, and the first batch of digitized visual materials has been made available to the general public through the e-Dunhuang digital library since May 2016. The aim of this study was to systematically evaluate e-Dunhuang from users’ perspectives, through usability testing with nine user tasks in different complexity levels and in-depth interviews with regard to a set of criteria in user experience. The results of quantitative analysis confirmed the overall effectiveness of e-Dunhuang in supporting user task completion and demonstrated significant improvements in several criteria over an earlier panorama collection of Dunhuang caves. The results of qualitative analysis revealed in-depth reasons for why participants felt satisfied with some criteria but had concerns with other criteria. Based on the findings, suggestions are proposed for further improvement in e-Dunhuang. As e-Dunhuang is a representative repository of digitized visual materials of cultural heritage, this study offers insights and empirical findings on user-centered evaluation of cultural heritage digital libraries.
• Hu, X., Cheong, C. W., & Chu, S. K. (2018). Developing a multidimensional framework for analyzing student comments in Wikis. Educational Technology and Society, 21(Issue 4).
  More info

  This study develops a framework for analyzing student comments in Wikis of group writing to inform learning assessment. It first drew on the literature to develop a framework consisting of three modules measuring student interaction, meaning construction and thinking development in the writing process. In-service teachers were interviewed to ensure framework practicality and inform subsequent refinement. A sample of 1,482 Wiki page comments was collected from 48 groups of secondary school students in Hong Kong to test the developed framework. Statistical analyses and association rule mining were conducted to the coded data to explore the relations among coding categories. This study aims to raise the attention on page comments in the analysis of student activities in Wiki and provided empirical evidence on category relations, which will be instructive for further research and practice in Wiki-supported learning.
• Hu, X., Cheong, C. W., Zhang, S., & Downie, J. S. (2018). Mood metadata on Chinese music websites: an exploratory study with user feedback. Online Information Review, 42(Issue 6). doi:10.1108/oir-01-2017-0023
  More info

  Purpose: Music mood is an important metadata type on online music repositories and stream music services worldwide. Many existing studies on mood metadata have focused on music websites and services in the Western world to the exclusion of those serving users in other cultures. The purpose of this paper is to bridge this gap by exploring mood labels on influential Chinese music websites. Design/methodology/approach: Mood labels and the associated song titles were collected from six Chinese music websites, and analyzed in relation to mood models and findings in the literature. An online music listening test was conducted to solicit users’ feedback on the mood labels on two popular Chinese music websites. Mood label selections on 30 songs from 64 Chinese listeners were collected and compared to those given by the two websites. Findings: Mood labels, although extensively employed on Chinese music websites, may be insufficient in meeting listeners’ needs. More mood labels of high arousal semantics are needed. Song languages and user familiarity to the songs show influence on users’ selection of mood labels given by the websites. Practical implications: Suggestions are proposed for future development of mood metadata and mood-enabled user interfaces in the context of global online music access. Originality/value: This paper provides insights on understanding the mood metadata on Chinese music websites and uniquely contributes to existing knowledge of culturally diversified music access.
• Hu, X., Ng, J., & Xia, S. (2018). User-Centered evaluation of metadata schema for nonmovable cultural heritage: Murals and stone cave temples. Journal of the Association for Information Science and Technology, 69(Issue 12). doi:10.1002/asi.24065
  More info

  Digitization provides a solution for documentation and preservation of nonmovable cultural heritages. Despite efforts for the preservation of cultural heritages around the world, no well-accepted metadata schema has been developed for murals and stone cave temples, which are often high-value heritages built in ancient times. In addition, the literature is scarce on the user-centered evaluation of metadata schemas of this kind. This study therefore aims to offer insights on developing and evaluating a metadata schema for organizing information of these historic and complex cultural heritages. In-depth interviews were conducted with a total of 30 users, including 18 professional and 12 public users, and interview transcripts were coded through a qualitative content analysis approach. Findings reveal the importance of specific metadata elements as perceived by the two groups of end users, which correlated with their cultural heritage information-seeking behaviors. In addition, the issues of standardization of cataloging of cultural heritage information and interoperability among metadata schemas have been raised by users for enhancing the user experience with digital platforms of cultural heritage information. The coding schema developed in this study can serve as a framework for follow-up evaluations of metadata schemas, contributing to the ongoing development of cultural heritage metadata.
• Lai, C., Hu, X., & Lyu, B. (2018). Understanding the nature of learners’ out-of-class language learning experience with technology. Computer Assisted Language Learning, 31(Issue 1-2). doi:10.1080/09588221.2017.1391293
  More info

  Out-of-class learning with technology comprises an essential context of second language development. Understanding the nature of out-of-class language learning with technology is the initial step towards safeguarding its quality. This study examined the types of learning experiences that language learners engaged in outside the classroom and the influencing factors. Three distinct types of technological experiences, with different incentives and different emotional and behavioral manifestations, were identified based on the interview responses of 21 university foreign language learners. Structural equation modeling analysis of 439 survey responses indicated that the three types of technological experiences were influenced differently by various attitudinal and support factors. Instruction-oriented technological experiences were influenced the most by learners’ perception of the usefulness of the technological experience for language learning, and entertainment- and information-oriented technological experiences were the only technological experiences that were influenced directly by perceived ease of the technological experience for language learning. Social-oriented technological experiences were influenced by myriad factors. Furthermore, it was found that the influencing factors for these experiences varied for learners with beginning and with intermediate proficiency levels. The findings underscore the importance of adopting differentiated approaches to supporting different types of technological experiences.
• Liu, M., Hu, X., & Li, J. (2018). Knowledge flow in China’s humanities and social sciences. Quality and Quantity, 52(Issue 2). doi:10.1007/s11135-017-0539-y
  More info

  Despite fruitful studies on knowledge flow and interdisciplinarity, there are few investigations on knowledge flow in humanities and social sciences (HSS) and how knowledge from science and technology diffuses to HSS sub-disciplines. Based on Chinese and English articles in HSS, this study explored knowledge flow in China’s HSS with an analysis of Chinese and English publications from 1998 to 2014. Findings include: (1) the interdisciplinarity degree of knowledge absorption in social sciences is higher than that of humanities in both Chinese and English articles, meanwhile the degree of interdisciplinarity in all HSS sub-disciplines increased constantly; (2) Chinese scholars in HSS increasingly tended to learn knowledge in hard sciences and applied it to their domains, especially in English articles; (3) in Chinese articles, Economics was the most crucial knowledge base, while Management, Education and Law were absorption-oriented sub-disciplines; in English articles Management, Law, Literature and Philosophy were absorption-oriented sub-disciplines.
• Liu, M., Hu, X., & Schedl, M. (2018). The relation of culture, socio-economics, and friendship to music preferences: A large-scale, cross-country study. PLoS ONE, 13(Issue 12). doi:10.1371/journal.pone.0208186
  More info

  Music listening is an inherently cultural behavior, which may be shaped by users’ backgrounds and contextual characteristics. Due to geographical, socio-economic, linguistic, and cultural factors as well as friendship networks, users in different countries may have different music preferences. Investigating cultural-socio-economic factors that might be associated with between-country differences in music preferences can facilitate music information retrieval, contribute to the prediction of users’ music preferences, and improve music recommendation in cross-country contexts. However, previous literature provides limited empirical evidence of the relationships between possible cross-country differences on a wide range of socio-economic aspects and those in music preferences. To bridge this research gap, and drawing on a large-scale dataset, LFM-1b, this study examines the possible relationship between cross-country differences in artist, album, and genre listening frequencies as well as the cross-country distance in geographical, socio-economic, linguistic, cultural, and friendship connections using the Quadratic Assignment Procedure. Results indicate: (1) there is no significant relationship between geographical and economic distance on album, artist, and genre preferences’ distance at the country-level; (2) the cross-country distance of three cultural dimensions (masculinity, long-term orientation, and indulgence) is positively associated with both the album and artist preferences distances; (3) the between-country distance in main languages has a positive relationship with the album, artist, and genre preferences distances across countries; (4) the density of friendship connections among countries negatively correlates to the cross-country preference distances in terms of artist and genre. Findings from this study not only expand knowledge of factors related to music preferences at the country level, but also can be integrated into real-world music recommendation systems that consider country-level music preferences.
• Liu, M., Hu, X., Wang, Y., & Shi, D. (2018). Survive or perish: Investigating the life cycle of academic journals from 1950 to 2013 using survival analysis methods. Journal of Informetrics, 12(Issue 1). doi:10.1016/j.joi.2018.02.001
  More info

  Since the emergence of the world's first academic journal in 1665, numerous academic journals have been launched and ceased publication. At the turn of the twenty-first century, academic journals are experiencing a dramatic revolution amidst increasingly fierce competition. However, limited research has investigated the survival pattern and the reasons why some academic journals have survived and others have not. Drawing on the data of academic journals in Ulrich's Periodicals Directory from 1950 to 2013, this study examined the life cycle of academic journals and revealed contributing factors related to the survival probabilities of academic journals using a Kaplan-Meier estimator, log-rank statistics, Cox proportional hazards models and propensity score matching. The results show that (1) the average survival rate of all the academic journals presents a rising-decreasing-rising pattern; (2) the third year after commencement is a peak year for academic journals to cease publication; (3) academic journals published in the UK, China, India and Russia, those in the field of technology, and those published in a single language cease publication sooner than their counterparts; (4) academic journals that provide online formats at launch time have a higher probability of surviving than non-online ones and those that provide online formats after launch time; (5) academic journals that provide print versions at launch time are more likely to survive than those without print formats and those that provide print formats after launch time; (6) academic journals that have a peer-reviewed process and that are published in multiple languages have a higher chance of survival; (7) academic journals published in English in China and Japan suffer a higher risk of termination than those published in native languages; (8) academic journals in the field of technology are more likely to cease publication than journals in the field of natural science; and (9) academic journals published in China can survive with a relatively high probability.
• Haruna, H., Tshuma, N., & Hu, X. (2017). Health Information Needs and Reliability of Sources Among Nondegree Health Sciences Students: A Prerequisite for Designing eHealth Literacy. Annals of Global Health, 83(Issue 2). doi:10.1016/j.aogh.2017.03.516
  More info

  Background Understanding health information needs and health-seeking behavior is a prerequisite for developing an electronic health information literacy (EHIL) or eHealth literacy program for nondegree health sciences students. At present, interest in researching health information needs and reliable sources paradigms has gained momentum in many countries. However, most studies focus on health professionals and students in higher education institutions. Objective The present study was aimed at providing new insight and filling the existing gap by examining health information needs and reliability of sources among nondegree health sciences students in Tanzania. Method A cross-sectional study was conducted in 15 conveniently selected health training institutions, where 403 health sciences students were participated. Thirty health sciences students were both purposely and conveniently chosen from each health-training institution. The selected students were pursuing nursing and midwifery, clinical medicine, dentistry, environmental health sciences, pharmacy, and medical laboratory sciences courses. Involved students were either in their first year, second year, or third year of study. Results Health sciences students' health information needs focus on their educational requirements, clinical practice, and personal information. They use print, human, and electronic health information. They lack eHealth research skills in navigating health information resources and have insufficient facilities for accessing eHealth information, a lack of specialists in health information, high costs for subscription electronic information, and unawareness of the availability of free Internet and other online health-related databases. Conclusion This study found that nondegree health sciences students have limited skills in EHIL. Thus, designing and incorporating EHIL skills programs into the curriculum of nondegree health sciences students is vital. EHIL is a requirement common to all health settings, learning environments, and levels of study. Our future intention is to design EHIL to support nondegree health sciences students to retrieve and use available health information resources on the Internet.
• Hu, X. (2017). Automated recognition of thinking orders in secondary school student writings. Learning: Research and Practice, 3(Issue 1). doi:10.1080/23735082.2017.1284253
  More info

  Despite the rapid development in the area of learning analytics (LA), there is comparatively little focused towards the secondary level of education. This ongoing work presents the latest developed function of Wikiglass, an LA tool designed for automatically recognising, aggregating, and visualising levels of thinking orders in student collaborative writing. Three levels of thinking orders were defined based on frameworks in writing assessment and an adapted Bloom’s taxonomy of cognitive domains. Text categorisation models were constructed and evaluated using machine learning and natural language processing techniques. The best performing model was then integrated into Wikiglass, with visualisations in different modes and scopes (i.e., class, group, individual). Currently being used in classrooms, Wikiglass is expected to assist teachers in identifying at-risk individuals and groups, refining assessment rubrics, and selecting example sentences as teaching materials, as well as to facilitate students in self-monitoring and reflection.
• Hu, X., & Kando, N. (2017). Task complexity and difficulty in music information retrieval. Journal of the Association for Information Science and Technology, 68(Issue 7). doi:10.1002/asi.23803
  More info

  There has been little research on task complexity and difficulty in music information retrieval (MIR), whereas many studies in the text retrieval domain have found that task complexity and difficulty have significant effects on user effectiveness. This study aimed to bridge the gap by exploring i) the relationship between task complexity and difficulty; ii) factors affecting task difficulty; and iii) the relationship between task difficulty, task complexity, and user search behaviors in MIR. An empirical user experiment was conducted with 51 participants and a novel MIR system. The participants searched for 6 topics across 3 complexity levels. The results revealed that i) perceived task difficulty in music search is influenced by task complexity, user background, system affordances, and task uncertainty and enjoyability; and ii) perceived task difficulty in MIR is significantly correlated with effectiveness metrics such as the number of songs found, number of clicks, and task completion time. The findings have implications for the design of music search tasks (in research) or use cases (in system development) as well as future MIR systems that can detect task difficulty based on user effectiveness metrics.
• Hu, X., & Yang, Y. H. (2017). Cross-Dataset and Cross-Cultural Music Mood Prediction: A Case on Western and Chinese Pop Songs. IEEE Transactions on Affective Computing, 8(Issue 2). doi:10.1109/taffc.2016.2523503
  More info

  In music mood prediction, regression models are built to predict values on several mood-representing dimensions such as valence (level of pleasure) and arousal (level of energy). Many studies have shown that music mood is generally predictable based on music acoustic features, but these experiments were mostly conducted on datasets with homogeneous music. Little research has been done to explore the generalizability of mood regression models cross datasets, especially those with music in different cultures. In the increasingly global market of music listening, generalizable models are highly desirable for automated processing, searching and managing music collections with heterogeneous characteristics. In this study, we evaluated mood regression models built on fifteen acoustic features in five mood-related musical aspects, with a focus on cross-dataset generalizability. Specifically, three distinct datasets were involved in a series of five experiments to examine the effects of dataset size, reliability of annotations and cultural backgrounds of music and annotators on mood regression performances and model generalizability. The results reveal that the size of the training dataset and the annotation reliability of the testing dataset affect mood regression performances. When both factors are controlled, regression models are generalizable between datasets sharing a common cultural background of music or annotators.
• Hu, X., Choi, K., & Downie, J. S. (2017). A framework for evaluating multimodal music mood classification. Journal of the Association for Information Science and Technology, 68(Issue 2). doi:10.1002/asi.23649
  More info

  This research proposes a framework for music mood classification that uses multiple and complementary information sources, namely, music audio, lyric text, and social tags associated with music pieces. This article presents the framework and a thorough evaluation of each of its components. Experimental results on a large data set of 18 mood categories show that combining lyrics and audio significantly outperformed systems using audio-only features. Automatic feature selection techniques were further proved to have reduced feature space. In addition, the examination of learning curves shows that the hybrid systems using lyrics and audio needed fewer training samples and shorter audio clips to achieve the same or better classification accuracies than systems using lyrics or audio singularly. Last but not least, performance comparisons reveal the relative importance of audio and lyric features across mood categories.
• Hu, X., Ho, E. M., & Qiao, C. (2017). Digitizing Dunhuang Cultural Heritage: A User Evaluation of Mogao Cave Panorama Digital Library. Journal of Data and Information Science, 2(Issue 3). doi:10.1515/jdis-2017-0014
  More info

  Purpose: This study is a user evaluation on the usability of the Mogao Cave Panorama Digital Library (DL), aiming to measure its effectiveness from the users' perspective and to propose suggestions for improvement. Design/methodology/approach: Usability tests were conducted based on a framework of evaluation criteria and a set of information seeking tasks designed for the Dunhuang cultural heritage, and interviews were conducted for soliciting in-depth opinions from participants. Findings: The results of the usability tests indicate that the DL was more efficient in supporting simple information seeking tasks than those of higher-complexity levels. Statistical tests reveal that there were correlations among dimensions of usability criteria and user effectiveness measures. Moreover, interview discourses exposed specific usability issues of the DL. Research limitations: This research is based on a relatively small sample size, resulting in a limited representativeness of user diversity. A larger sample size is needed for a systematic cross group comparison. Practical implications: This study evaluated the usability of the Mogao Cave Panorama DL and proposed suggestions for its improvement for better experience. The results also provide a reference to other cultural heritage DLs with panorama functions. Originality/value: This study is one of the first evaluating cultural heritage DLs from the perspective of user experience. It provides methodological references for relevant studies: the evaluation framework, the designed information seeking tasks, and the interview questions can be adopted or adapted in evaluating other visually centric DLs of cultural heritage.
• Hu, X., Ng, J., Xia, S., & Fu, Y. K. (2017). Evaluating metadata schema for murals and stone cave temples: Towards digitizing cultural heritage. Proceedings of the Association for Information Science and Technology, 54(Issue 1). doi:10.1002/pra2.2017.14505401054
  More info

  Despite efforts and attempts on preserving cultural heritages around the world, few studies have been conducted on metadata schema of murals and stone cave temples which are often high value heritages dating from ancient times. The literature is also scarce in user-centered evaluation of metadata schema of this kind. This study therefore aims to offer insights on developing and evaluating a drafted metadata schema for digitizing and preserving historic murals and stone cave temples. In-depth interviews were conducted with 15 scholars and professionals whose work was related to murals and stone cave temples. A coding framework was designed to analyze the interview transcripts. Initial findings show that users' seeking behaviors for cultural heritage information corroborated with their perceived importance of specific metadata elements. The coding schema developed in this study can serve as a framework for follow-up evaluations of metadata schemas of this kind, contributing to the ongoing development of cultural heritage metadata.
• Hu, X., Yu, B., Alman, S., Renear, A. H., & Carbo, T. (2017). Teaching information science and technology to the world? Practices, challenges and visions. Proceedings of the Association for Information Science and Technology, 54(Issue 1). doi:10.1002/pra2.2017.14505401074
  More info

  Information science and technology (IST) has made unprecedented impacts on society. Teaching IST should by no means be limited within the boundary of the discipline. Based on their experiences in teaching IST in various levels, speakers in this panel will reflect the past, share current practices and envision the future of teaching IST to the world. Collectively we aim to stimulate further discussions on increasing the impact of the field through education.
• Tavernier, M., & Hu, X. (2017). Developing young children's technology-based communication skills using iPads and creation apps: An action study. Proceedings of the Association for Information Science and Technology, 54(Issue 1). doi:10.1002/pra2.2017.14505401045
  More info

  Today's young children are exposed to new information technology and communication practices from early on, but the usage activities they engage in may not help fulfilling their needs in developing technology-based communication skills. This article presents preliminary findings of an action study with eight children of four to five years old in using one creation app, SeeSaw, to create artifacts to express themselves in ways that suit their personal abilities. A sequence of 11 teacher-initiated activities guided the children towards the development of age appropriate technology-based communication skills (TBCS). Digital artifacts produced by the children and videos of learning sessions were collected and analyzed using a thematic content analysis approach. The findings reveal that four- to five-year-old children acquire TBCS quickly and progressively. They are also able to use their developing TBCS effectively to communicate their experiences, knowledge and ideas using a variety of audio and visual elements. This study contributes to the literature of new learning opportunities in this information environment, with a focus on the user group of young children which has been understudied in information science.
• Aytac, S., Ma, L., Potnis, D., Rorissa, A., Chen, H. L., & Hu, X. (2016). Diversity and multiculturalism of LIS education. Proceedings of the Association for Information Science and Technology, 53(Issue 1). doi:10.1002/pra2.2016.14505301003
  More info

  The purpose of the panel is to raise the common questions about diversity and multiculturalism training in library schools, and engage the audience in a meaningful discussion about diversity and multiculturalism. We will discuss the importance of diversity and multiculturalism training in library and information science curricula by examining multiple countries/regions as case studies. We would be seeking answers to two major questions: “What do responsible library science educators need to do to educate librarians on diversity/multiculturalism/internationalism?” and “How can we educate future librarians so that they will have a greater perspective on diversity and multiculturalism?”.
• Choi, K., Lee, J. H., Hu, X., & Downie, J. S. (2016). Music subject classification based on lyrics and user interpretations. Proceedings of the Association for Information Science and Technology, 53(Issue 1). doi:10.1002/pra2.2016.14505301041
  More info

  That music seekers consider song subject metadata to be helpful in their searching/browsing experience has been noted in prior published research. In an effort to develop a subject-based tagging system, we explored the creation of automatically generated song subject classifications. Our classifications were derived from two different sources of song-related text: 1) lyrics; and 2) user interpretations of lyrics collected from songmeanings.com. While both sources contain subject-related information, we found that user-generated interpretations always outperformed lyrics in terms of classification accuracy. This suggests that user interpretations are more useful in the subject classification task than lyrics because the semantically ambiguous poetic nature of lyrics tends to confuse classifiers. An examination of top-ranked terms and confusion matrices supported our contention that users' interpretations work better for detecting the meaning of songs than what is conveyed through lyrics.
• Hu, X., & Lee, J. H. (2016). Towards global music digital libraries: A cross-cultural comparison on the mood of Chinese music. Journal of Documentation, 72(Issue 5). doi:10.1108/jd-01-2016-0005
  More info

  Purpose: The purpose of this paper is to compare music mood perceptions of people with diverse cultural backgrounds when they interact with Chinese music. It also discusses how the results can inform the design of global music digital libraries (MDL). Design/methodology/approach: An online survey was designed based on the Music Information Retrieval Evaluation eXchange (MIREX) five-cluster mood model, to solicit mood perceptions of listeners in Hong Kong and the USA on a diverse set of Chinese music. Statistical analysis was applied to compare responses from the two user groups, with consideration of different music types and characteristics of listeners. Listeners’ textual responses were also analyzed with content coding. Findings: Listeners from the two cultural groups made different mood judgments on all but one type of Chinese music. Hong Kong listeners reached higher levels of agreement on mood judgments than their US counterparts. Gender, age and familiarity with the songs were related to listeners’ mood judgment to some extent. Practical implications: The MIREX five-cluster model may not be sufficient for representing the mood of Chinese music. Refinements are suggested. MDL are recommended to differentiate tags given by users from different cultural groups, and to differentiate music types when classifying or recommending Chinese music by mood. Originality/value: It is the first study on cross-cultural access to Chinese music in MDL. Methods and the refined mood model can be applied to cross-cultural access to other music types and information objects.
• Huang, C., Zhao, J. H., & Hu, X. (2005). Sustainable development OCR system in CADAL application. Journal of Zhejiang University: Science, 6(Issue 11). doi:10.1631/jzus.2005.a1312
  More info

  This paper briefly introduces the main ideas of a sustainable development OCR system based on open architecture techniques and then describes the construction of an optical character recognition (OCR) center built on computer clusters, for the purpose of dynamically improving the recognition precision of the digitized texts of a million volumes of books produced by the China-US Million Books Digital Library (CADAL) Project. The practice of this center will provide helpful reference for other digital library projects.

Proceedings Publications

• Hernández López, N. P., Hu, X., & Ng, D. T. (2025). Insights from Culturally Relevant AI Education Programme for Secondary School Students. In International Conference on Artificial Intelligence in Education.
  More info

  Artificial Intelligence (AI) continues to gain relevance, and with it, the need to initiate students to the knowledge and use of AI technology. Despite numerous recent attempts at facilitating AI education for young learners, there is still a lack of research that leverages students’ cultural identity for improving learning about AI. Therefore, we present a study aimed at investigating the effect of Culturally Relevant Artificial Intelligence Education (CRAIEd) over students’ learning outcomes, including behavioural, cognitive, and affective. We focus on the use of music as a cultural signifier that can connect with students’ identity and motivate learning. Through a mixed-methods approach, we analyse tests, self-reported questionnaires, and students’ self-reflections, to evaluate the influence of CRAIEd on students’ knowledge and attitudes towards AI. Findings reveal that CRAIEd that explicitly involves students’ cultural background has a positive impact over students’ learning outcomes. Implications from this study will contribute to the understanding of learning about AI in a culturally relevant educational programme, and at the same time having practical implications for fostering learning about AI situated in a specific cultural context.
• Lau, K. W., Cheng, L., & Hu, X. (2025). Empowering Primary School Students to Create Virtual Reality Content: An Outreach Model for Digital Libraries. In 24th ACM/IEEE Joint Conference on Digital Libraries, JCDL 2024.
  More info

  Digital maker activities can enhance learners agency, cultivating their digital and multiliteracy skills, and improving their creativity. The products from digital maker activities can potentially enrich the content of digital libraries and be shared with everyone with network access. Virtual reality (VR) content, owing to its affordance in providing audiences with immersive experiences, is getting popular in education and information services. The integration of digital maker activity and VR, namely VR content creation, can potentially provide benefits from the two, yet can have a high learning curve for the creators. In this study, we leverage a low-Tech VR content creation approach, to teach primary school students to create VR stories in the context of an environment conservation project. Preliminary results from 85 primary school students demonstrate the effectiveness of this approach in improving their digital literacy, which provides a viable model for outreach activities in digital libraries and related institutions.
• Lin, W., & Hu, X. (2025). GenAI-Supported Creative Learning in Digital Museum Education: A Case Study of Maritime Art Painting Creation. In 2025 CHI Conference on Human Factors in Computing Systems, CHI EA 2025.
  More info

  Creative learning has significantly enriched on-site museum education by fostering deeper engagement, encouraging exploration, and promoting active participation through hands-on experiences. However, the digital transformation of holistic and structured creative learning processes in museum education remains largely underexplored. Although Generative Artificial Intelligence (GenAI) holds promise for supporting creative learning, its full integration across all phases in non-formal education settings like museums remains limited. In this work, we designed Mariscope, a GenAI-supported creative learning platform for marine museum education. By integrating GenAI with the five stages of the iterative creative learning path, Mariscope offers a personalized and dynamic learning experience. A preliminary study with four participants demonstrated positive impacts on learning outcomes and creative self-efficacy, showcasing the platform’s potential for enhancing digital museum education through iterative and interactive creative learning processes.
• Lin, W., Hu, X., & Xu, J. (2025). Leveraging Augmented Reality and Tangible User Interface to Enhance Human-Information Interaction in Digital Libraries. In 24th ACM/IEEE Joint Conference on Digital Libraries, JCDL 2024.
  More info

  By integrating digital content with the physical world, augmented reality (AR) creates a seamless blend of digital information and real environments. Simultaneously, a tangible user interface (TUI) enables users to interact with digital information through physical objects, promoting more natural and effective content understanding and manipulation. In this preliminary study, we develop an interface, "Poetic Moon,"that combines AR and TUI for human-information interaction within the context of Chinese poetry in digital libraries. Our results indicate this approach offers immersive and intuitive experiences, enhancing users' engagement and understanding of the presented content. This method has the potential for digital libraries, offering innovative ways to explore and interact with content in literary and other subjects.
• Liu, Y., Ma, Z., Ng, J. T., & Hu, X. (2025). Multimodal learning analytics for game-based assessment of collaborative problem solving skills among young students. In 15th International Conference on Learning Analytics and Knowledge, LAK 2025.
  More info

  Collaborative Problem Solving (CPS) has emerged as a key competence for the 21st century. In support of this, valid assessments of CPS skills have become critical. However, limited research has designed and developed CPS assessments for young students. Based on multimodal learning analytics, we aim to develop and validate a game-based assessment of CPS for primary school students. In this study, evidence centered design approach was used to design and develop the game-based CPS assessment. Specifically, we designed and developed a mobile multiplayer online 3D role-playing game on CPS and a coding scheme for coding students' gameplay data (i.e., game logs and voice chat) based on the ATC21S CPS framework. A total of 32 primary 5 students participated in this study to play the game in a group of four and complete a questionnaire of CPS skills. The gameplay data were coded based on our coding scheme. Correlation analysis between the coded results and the CPS questionnaire data supported the criterion validity of our game-based assessment measure. Additionally, the results of expert interview facilitated our understanding of assessment design and data use. This study will make methodological and practical contributions to the integration of MMLA into game-based CPS assessments.
• Wang, Z., Lin, W., & Hu, X. (2025). Self-service Teacher-facing Learning Analytics Dashboard with Large Language Models. In 15th International Conference on Learning Analytics and Knowledge, LAK 2025.
  More info

  With the rise of online learning platforms, the need for effective learning analytics (LA) has become critical for teachers. However, the development of traditional LA dashboards often requires technical expertise and a certain level of data literacy, preventing many teachers from integrating LA dashboards effectively and flexibly into their teaching practice. This paper explores the development of a self-service teacher-facing learning analytics dashboard powered by large language models (LLMs), for improving teaching practices. By leveraging LLMs, the self-service system aims to simplify the implementation of data queries and visualizations, allowing teachers to create personalized LA dashboards using natural languages. This study also investigates the capabilities of LLMs in generating charts for LA dashboards and evaluates the effectiveness of the self-service system through usability tests with 15 teachers. Preliminary findings suggest that LLMs demonstrate high capabilities in generating charts for LA dashboards, and the LLM-powered self-service system can effectively address participating teachers' pedagogical needs for LA. This research contributes to the ongoing research on the intersection of LLMs and education, emphasizing the potential of self-service systems to empower teachers in daily teaching practices.
• Cheng, K. L., Chan, C. K., Tu, Y., & Hu, X. (2024). Examining Students' Online Learning and Collaboration Using Analytics-Supported Assessment Tools and Dashboards. In 24th IEEE International Conference on Advanced Learning Technologies, ICALT 2024.
  More info

  This paper reports on the preliminary findings of designing a learning analytical tool to assess and facilitate online collaborative learning in discussion forums. The analytics tool developed is grounded on collaboration theories to unravel students' online collaboration, encompassing three features: (a) participation and build-on posts, (b) lexical keywords for domain understanding, and (c) communicative acts for dialogic interactions. The tool was designed to analyze students' online discussions for two classes in a postgraduate educational studies course. Findings unravel student online collaboration behaviour, including (a) high engagement with online posts exceeding course requirements, (b) frequency/ links among keywords used (not used) indicate students' knowledge networks and gaps, and (3) dialogic communication acts commonly employed while higher-level acts (e.g., coordination) not yet adopted. Findings also indicate student groups using a higher frequency of communication acts (more dialogic in discussion) also obtained higher grades, providing some validation. Implications suggest how the tools can be used to assess social-semantic-dialogic online collaborative behaviour and how instructors can use analytics information to adapt their instructional strategies to address students' knowledge gaps and provide feedback. Students can also use the analytics information to regulate and improve online discussions.
• Hernández López, N. P., & Hu, X. (2024). Culturally Relevant Artificial Intelligence Education with Music for Secondary School Students. In International Conference on Artificial Intelligence in Education.
  More info

  Artificial Intelligence (AI) has increasingly gained attention in recent years, and with it, the need to involve youth in responsible uses the technology. I propose a method that aims to equip secondary school students with the necessary knowledge and skills to identify, describe, interact, and create with AI responsibly. Particularly, my research aims to engage students in culturally relevant learning integrating music as a cultural signifier. Using a Constructionist approach, I propose a method in which students will learn about AI through making personally meaningful musical artefacts. By situating the study in two different sociocultural contexts, this study aims to characterise the extent to which different local cultures influence learning outcomes (cognitive, affective, and behavioural) when learning about AI. This study aims to provide a systematic method for collecting and analysing data from students’ interactions with an AI-enabled music creation platform, thus enabling a holistic understanding of students’ learning processes.
• Liu, Y., Dong Ng, J. T., Hu, X., & Ma, Z. (2024). Towards Multimodal Learning Analytics of Game-based Collaborative Problem Solving among Primary School Students. In 24th IEEE International Conference on Advanced Learning Technologies, ICALT 2024.
  More info

  Well-designed digital games can serve as the vehicle to assess and support young people' collaborative problem solving (CPS) skills. However, there is limited research leveraging multimodal learning analytics (MmLA) to explore students' game-based CPS processes and outcomes. Inspired by MmLA methods and approaches, this preliminary study aims to examine students' demonstration of CPS skills through collecting and analyzing a dataset of combined game logs and verbal discourses from two groups of primary school students with contrasting performances. Based on the Assessment and Teaching of 21st Century Skills CPS framework, we iteratively coded the dataset. Results of descriptive statistics showed that the successful group exhibited cognitive skills more frequently while the unsuccessful group showcased social skills more. Results of epistemic network analysis (ENA) revealed that, in both social and cognitive dimensions, the successful group demonstrated more diverse and stronger associations among various subskills, whereas there were fewer associations in the unsuccessful group. Implications are drawn for MmLA and CPS research and teaching practices of CPS skills.
• Que, Y., Ng, J. T., Hu, X., Mak, M. K., & Yip, P. T. (2024). Using Multimodal Learning Analytics to Examine Learners' Responses to Different Types of Background Music during Reading Comprehension. In 14th International Conference on Learning Analytics and Knowledge, LAK 2024.
  More info

  Previous studies have evaluated the affordances and challenges of performing cognitively demanding learning tasks with background music (BGM), yet the effects of various types of BGM on learning still remain an open question. This study aimed to examine the impacts of different music genres and fine-grained music characteristics on learners' emotional, physiological, and pupillary responses during reading comprehension. Leveraging multimodal learning analytics (MmLA) methods of collecting data in multiple modalities from learners, a user experiment was conducted on 102 participants, with half of them reading with self-selected BGM (i.e., the experimental group), while the other half reading without BGM (i.e., the control group). Results of statistical analyses and interviews revealed significant differences between the two groups in their self-reported emotions and automatically measured physiological responses when the experimental group was exposed to classical, easy-listening, rebellious and rhythmic music. Fine-grained music characteristics (e.g., instrumentation, tempo) could predict learners' emotions, pupillary, and physiological responses during reading comprehension. The expected contributions of this study include: 1) providing empirical evidence for understanding affective dimensions of learning with BGM, 2) applying MmLA methods for examining the impacts of BGM on learning, and 3) yielding practical implications on how to improve learning with BGM.
• Que, Y., Zheng, Y., Hsiao, J. H., & Hu, X. (2024). Predicting Learners' Meta-cognition Using Eye Movements during Reading with Background Music. In 24th IEEE International Conference on Advanced Learning Technologies, ICALT 2024.
  More info

  Many students enjoy listening to background music (BGM) when they read, but it is challenging to measure their metacognitive states (e.g., understanding of the passage, engagement in reading). Eye movements, as an approach in multimodal learning analytics (MmLA), can offer continuous fine-grained data that reflect learners' cognitive processes. This study explored the potential of utilizing eye movement measures to predict learners' meta-cognition during reading with BGM. Results showed that learners' eye movement measures integrated with the characteristics of the BGM, learner traits, and text complexity could predict their meta-cognitive states in reading. Findings can advance our understanding of human meta-cognition in multi-channel learning settings and provide insights for personalized BGM recommendations to enhance reading experiences.
• Wang, C., Hu, X., Hernández López, N. P., & Ng, J. T. (2024). Needs Analysis of Learning Analytics Dashboard for College Teacher Online Professional Learning in an International Training Initiative for the Global South. In 14th International Conference on Learning Analytics and Knowledge, LAK 2024.
  More info

  Online courses enable wide access to educational resources and thus provide a feasible platform for cross-regional teacher professional learning. Learning analytics dashboards (LAD) can support online learners by providing fine-grained feedback generated from learners' interactions with platforms. Nevertheless, most studies on teacher online professional learning focus on resource-rich and technology-advanced regions, with scarce attention to the Global South. Furthermore, existing studies on LAD design mainly target students' learning, rather than teachers' professional learning. Therefore, it is much needed to develop LAD for teacher-learners online professional learning in the Global South. Contextualized in an international online professional training initiative, this study conducted in-depth interviews with 42 teacher-learners from 19 countries in the Global South, aiming to identify their needs for 1) support on their self-regulated learning (SRL), and 2) potential LA components in dashboards. Findings indicated that teacher-learners needed support for self-regulated learning strategies, including motivation maintenance, time management, environment structuring, help-seeking, and self-evaluation. Nine LA features were identified to design the LADs to support SRL preliminarily. This co-designed LAD study with interviewees improved our understanding on the needs of college teachers in the Global South for LA support during their online professional learning, generating practical insights into needs-driven LAD designs.
• Wang, C., Ng, J. T., López, N. P., & Hu, X. (2024). Preliminary Evaluation of Learning Analytics Dashboard for College Teachers' Online Professional Learning. In 24th IEEE International Conference on Advanced Learning Technologies, ICALT 2024.
  More info

  Widely accessible online courses provide a feasible platform for teachers' continuous online professional learning. The learning analytics dashboard (LAD) provides fine-grained and actionable feedback that supports learners' self-regulated learning. However, previous studies on LAD design and evaluation predominantly focused on student-facing LADs, with scarce attention on LADs designed for teacher-learners. This study introduces the LAD in an online learning platform for college teachers and conducts a preliminary evaluation with 18 participants. Results show their largely positive ratings on five criteria (e.g., perceived usefulness, ease of use, and behavioral changes) and offer feedback for further refinements of the LAD. This study will improve our understanding of LA-enabled teacher online professional learning and provide practical implications for designing and evaluating LA tools catered to teacher-learners.
• Wang, Z., Ng, J. T., & Hu, X. (2024). Learning Analytics for Collaboration Quality Assessment during Virtual Reality Content Creation. In 24th IEEE International Conference on Advanced Learning Technologies, ICALT 2024.
  More info

  In this paper, we present an empirical study on collaborative virtual reality (VR) maker activities. A platform called CLEVR was designed to facilitate real-time VR co-creation with group awareness features. We develop an assessment framework for collaborative quality using log-based learning analytics (LA). By conducting time-series analysis on five pairs of participants, we demonstrate the potential of our platform and methodology for the development of LA tools to understand and support collaborative learning.
• Wang, Z., Ng, J. T., Que, Y., & Hu, X. (2024). Unveiling Synchrony of Learners' Multimodal Data in Collaborative Maker Activities. In 14th International Conference on Learning Analytics and Knowledge, LAK 2024.
  More info

  While current evaluation of maker activities has rarely explored students' learning processes, the multi-perspective and multi-level nature of collaboration adds complexity to learning processes of collaborative maker activities. In terms of group dynamics as an important indicator of collaboration quality, extant studies have shown the benefits of synchrony between learners' actions during collaborative learning processes. However, synchrony of learners' cognitive processes and visual attention in collaborative maker activities remains under-explored. Leveraging the multimodal learning analytics (MMLA) approach, this pilot study examines learners' synchrony patterns from multiple modalities of data in the collaborative maker activity of virtual reality (VR) content creation. We conducted a user experiment with five pairs of students, and collected and analyzed their electroencephalography (EEG) signals, eye movement and system log data. Results showed that the five pairs of collaborators demonstrated diverse synchrony patterns. We also discovered that, while some groups exhibited synchrony in one modality of data before becoming not synchronized in another modality, other groups started with a lack of synchrony followed by maintaining synchrony. This study is expected to make methodological and practical contributions to MMLA research and assessment of collaborative maker activities.
• Zhang, C., Hu, X., & Wei, W. (2024). A Review of AI-based Techniques in Online Course Recommendation: Metrics, Factors, and Research Methods. In 3rd International Conference on Intelligent Education and Intelligent Research, IEIR 2024.
  More info

  This paper presents a comprehensive literature review of online course recommendation systems, focusing on studies published between 2019 and 2023 across various academic databases. We selected and analyzed relevant research, covering key areas such as AI methods, research questions, influencing factors, and evaluation techniques. AI methods for course recommendation are categorized as traditional and advanced approaches. Traditional methods include Content-Based (CB) recommendation, Collaborative Filtering (CF), and related techniques, while advanced methods encompass deep learning and reinforcement learning (RL) techniques. Advanced AI methods demonstrate the ability to manage more complex data structures and provide more accurate and personalized recommendations compared to traditional methods. The common research challenges identified include accuracy, cold start, data sparsity, information overload, dynamic interest recommendation, and interpretability. We further analyzed the factors considered in these systems by categorizing user and course attributes. User attributes are divided into five categories: Demographic Information, User Behavior, Personal Preferences, Knowledge and Skills, and Emotional State and Social Interaction. Course attributes are categorized into four types: Course Metadata, Course Evaluation, Knowledge and Skills, and Social Influence. The most commonly used evaluation metrics include Recall (R), Precision (P) and F1-Score (F), which are straightforward to interpret and facilitate comparison across different studies. Additionally, we highlight future research directions, providing insights for further academic exploration in this domain.
• Zhou, W., Hu, X., Que, Y., & Wei, W. (2024). Immersive Interface Design for Cultural Heritage Learning Experience: An Exploration with Multimodal Learning Analytics. In 24th IEEE International Conference on Advanced Learning Technologies, ICALT 2024.
  More info

  This study employed multimodal learning analytics methods to evaluate different virtual reality (VR) interfaces, exploring their impact on the learning experience of cultural heritage. We recorded and analyzed the participants' learning behavior, self-reported perceptions from questionnaires' responses, electroencephalogram (EEG) signals, and visual fatigue. Preliminary results suggested that VR interfaces with more modules per scene led to higher efficiency of use, although more modules did not improve viewing experience. This study provides insights for immersive interface design in cultural heritage learning.
• Liu, R., Wang, Z., Ba, S., & Hu, X. (2023). Preliminary Exploration of the Effectiveness of Music Listening and Music Recommender for Studying in Naturalistic Settings. In 23rd IEEE International Conference on Advanced Learning Technologies, ICALT 2023.
  More info

  Listening to music is a common behavior when people study or work. However, effects of music listening on studying are still disputed in previous studies. To explore the associations between music characteristics and learning performance and engagement and to develop a music recommender for studying in naturalistic settings, we conducted a two-month field experiment with 51 undergraduate and graduate students. A mobile application based on the experience sampling method was designed and implemented to ubiquitously collect participants' learning status and music listening traces. Statistical tests and machine learning were adopted respectively for uncovering the associations between music listening on learning and constructing a music recommendation model. Results first indicated that learners' music preferences and several musical features were positively correlated with self-reported learning performance and concentration. Furthermore, machine learning modeling demonstrated promising results for developing a music recommender for studying in naturalistic settings. Findings are expected to contribute to research on learning with background music and learning-oriented music recommendation.
• López, N. P., & Hu, X. (2023). What Can Students Learn From Their Own Data? Data Literacy With Student-Facing Learning Analytics. In 17th International Conference of the Learning Sciences, ICLS 2023.
  More info

  With the increasing need to make sense of the ever-growing quantity of data originated from digital interactions, data literacy skills become a basic requirement to navigate everyday tasks. In the field of education, data has gained wide attention, especially with the introduction of analytics from teaching and learning data. Current trends of research on data literacy in learning sciences focus on educators' needs of specific training and knowledge about how to make data-driven decisions that benefit students' progress. Despite little research at the intersection of developing learning analytics (LA) for students and developing their data literacy skills, we argue that student-facing learning analytics (SFLA) can be leveraged for strengthening students' data knowledge and skills. Based on an integrative review of existing literature, we briefly discuss several important considerations that will benefit future implementations at the intersection of SFLA and data literacy.
• Ng, J. T., Liu, R., Wang, Z., & Hu, X. (2023). Automated Analysis of Text in Student-Created Virtual Reality Content. In 23rd IEEE International Conference on Advanced Learning Technologies, ICALT 2023.
  More info

  Assessments of digital maker activities increasingly rely on automatically analyzing student-created products and their components, such as their textual output. In particular, recent learning analytics research has proposed incorporating text analytic feedback for facilitating students' virtual reality (VR) content creation, though lacking direct empirical evidence from student-created artefacts. Thus, this study examined the relationships between metrics on text in student-created VR content and their learning performance. VR narration scripts and performance scores were collected from 102 students in a maker-based general education course. Results of statistical testing and text mining show that high and low-performing students demonstrated significant differences in such metrics as word counts, vocabulary sizes, and frequent unigrams and bigrams. This study makes methodological and practical contributions in the domains of maker education and learning analytics.
• Ng, J. T., Liu, Y., Chui, D. S., Man, J. C., & Hu, X. (2023). Leveraging LMS Logs to Analyze Self-Regulated Learning Behaviors in a Maker-based Course. In 13th International Conference on Learning Analytics and Knowledge: Towards Trustworthy Learning Analytics, LAK 2023.
  More info

  Existing learning analytics (LA) studies on self-regulated learning (SRL) have rarely focused on maker education that emphasizes student autonomy in their learning process. Towards using LA methods for generating evidence of SRL in maker-based courses, this study leverages logs of a learning management system (LMS) with its activity design aligned with the maker-based pedagogy. We explored frequencies and sequential patterns of students' SRL behaviors as reflected in the LMS logs and their relations with learning performance. Adopting a mixed method approach, we collected and triangulated both quantitative (i.e., system logs, performance scores) and qualitative (i.e., student-written reflections) data sources from 104 students. Based on current LA-based SRL research, we developed an LMS log-based analytic framework to define the SRL phases and behaviors applicable to maker activities. Statistical, data mining, and qualitative analysis methods were conducted on 48,602 logged events and 131 excerpts extracted from student reflections. Results reveal that high-performing students demonstrated some SRL behaviors (e.g., Making Personal Plans, Evaluation) more frequently than their low-performing counterparts, yet the two groups showcased fairly similar sequences of SRL behaviors. Theoretical, methodological and pedagogical implications are drawn for LA-based SRL research and maker education.
• Wang, C., Ba, S., Hu, X., & Shao, Y. (2023). Exploring Factors Limiting Participation in an Online Training Program for College Teachers from Developing Countries. In 23rd IEEE International Conference on Advanced Learning Technologies, ICALT 2023.
  More info

  Online video courses allow large-scale distribution of educational resources and thus provide a feasible platform for cross-regional teacher professional development (TPD). However, with most studies focusing on TPD using online video courses in developed countries, less attention was paid to factors influencing the online learning experience of teachers from less developed areas. Therefore, this pilot study aims at exploring predictors of 3471 college teachers' online learning engagement. First, the results of a multivariate linear regression (MLR) revealed that at the individual level, there was a significant positive relationship between teachers' age and learning duration. At the institutional level, participants from the partner institutions had longer learning durations. At the country level, there was a significant positive relationship between country literacy rates and learning durations. Then, another MLR on relationships between self-reported learning perceptions after taking the course and learning durations was conducted. Results showed that learners' prior knowledge was negatively associated with learning durations. There was a significantly positive relationship between 'recommend this platform to others' and learning duration. The findings can inform us of the factors that may have been overlooked when supporting college teachers' online professional learning in developing countries.
• Ba, S., Hu, X., Kong, R., & Law, N. (2022). Supporting adolescents' digital well-being in the post-pandemic era: Preliminary results from a multimodal learning analytics approach. In 22nd International Conference on Advanced Learning Technologies, ICALT 2022.
  More info

  Affected by the Covid-19 pandemic, the way adolescents receive their education has changed drastically from offline classrooms to online digital space. Despite the benefits of digital devices, we must also be cautious of the possible negative impacts of using digital devices excessively. In this study, we proposed a smart planning course to support adolescents in managing daily digital device usage. Meanwhile, we examined the effects of this course through a novel multimodal learning analytics (MMLA) approach. Although results of the quasi-experiment indicated few significant effects of the intervention, possibly due to its timing, the proposed MMLA approach was shown to provide more comprehensive and refined data compared to traditional methods. Future studies can use this approach for further activity-based analysis of students' digital well-being.
• Hu, X., Ng, J. T., Lee, C. N., & Tsang, H. Y. (2022). Towards Automated Analysis of Undergraduate Academic Writing using Metadiscourse, Cognitive Level and Word Network. In 16th International Conference of the Learning Sciences, ICLS 2022.
  More info

  Automated assessment of students' academic writing can provide timely feedback and alleviate teachers' workload. This study examined 199 undergraduate students' essays through metrics of metadiscourse, cognitive levels, and word network, and the relationships between these metrics and performances. Metrics were calculated under the framework of Hyland's metadiscourse model and Bloom Taxonomy whereas relationships were revealed by correlation analysis and classification with multiple algorithms. Findings show that students employed more interactive metadiscourse markers than interactional ones, and there were weak to mild positive correlations between some of these metrics and performance score. High-performing essays involved more higher-order thinking and stronger connections among words. Besides, classification models combining all three types of metrics were most effective in differentiating high and low-performing essays. The three types of metrics evaluated can be potentially used to provide fine-grained suggestions for improving students' academic writing.
• Ng, J. T., Sun, Y., & Hu, X. (2022). An Exploratory Study on Students' Digital Curation Competency and Experience in a General Education Course. In 16th International Conference of the Learning Sciences, ICLS 2022.
  More info

  Essential in this information landscape, digital curation is broadly defined as the practices of digitally organizing and re-presenting the cultural record of humankind for creating value and impact, producing visual and textual information for presenting a narrative to audiences. Despite the prevalence of our daily-life curatorial practices and being well-suited to a diversity of disciplines, training in digital curation knowledge and skills has not been democratized to be inclusive for students from non-technical backgrounds. This pilot study 1) proposes a maker activity, namely creating digital galleries of cultural heritage, in a general education course, 2) examines to what extent this learning activity helps develop digital curation competency of undergraduate students, and 3) investigates how this digital curation experience benefits their professional and personal pursuits.
• Tzi-Dong Ng, J., Hu, X., & Que, Y. (2022). Towards Multi-modal Evaluation of Eye-tracked Virtual Heritage Environment. In 12th International Conference on Learning Analytics and Knowledge: Learning Analytics for Transition, Disruption and Social Change, LAK 2022.
  More info

  In times of pandemic-induced challenges, virtual reality (VR) allows audience to learn about cultural heritage sites without temporal and spatial constraints. The design of VR content is largely determined by professionals, while evaluations of content often rely on learners' self-report data. Learners' attentional focus and understanding of VR content might be affected by the presence or absence of different multimedia elements including text and audio-visuals. It remains an open question which design variations are more conducive for learning about heritage sites. Leveraging eye-tracking, a technology often adopted in recent multimodal learning analytics (MmLA) research, we conducted an experiment to collect and analyze 40 learners' eye movement and self-reported data. Results of statistical tests and heatmap elicitation interviews indicate that 1) text in the VR environment helped learners better understand the presented heritage sites, regardless of having audio narration or not, 2) text diverted learners' attention away from other visual elements that contextualized the heritage sites, 3) exclusively having audio narration best simulated the experience of a real-world heritage tour, 4) narration accompanying text prompted learners to read the text faster. We make recommendations for improving the design of VR learning materials and discuss the implications for MmLA research.
• Tzi-Dong Ng, J., Wang, Z., & Hu, X. (2022). Needs Analysis and Prototype Evaluation of Student-facing la Dashboard for Virtual Reality Content Creation. In 12th International Conference on Learning Analytics and Knowledge: Learning Analytics for Transition, Disruption and Social Change, LAK 2022.
  More info

  Being a promising constructionist pedagogy in recent years, maker education empowers students to take agency of their learning process through constructing both knowledge and real-world physical or digital products and fosters peer interactions for collective innovation. Learning Analytics (LA) excels at generating personalized, fine-grained feedback in near real-time and holds much potential in supporting process-oriented and peer-supported learning activities, including maker activities. In the context of virtual reality (VR) content creation for cultural heritage education, this study qualitatively solicited 27 students' needs on progress monitoring, reflection, and feedback during their making process. Findings have inspired the prototype design of a student-facing LA dashboard (LAVR). Leveraging multimodal learning analytics (MmLA) such as text and audio analytics to fulfill students' needs, the prototype has various features and functions including automatic task reminders, content quality detection, and real-time feedback on quality of audio-visual elements. A preliminary evaluation of the prototype with 10 students confirms its potential in supporting students' self-regulated learning during the making process and for improving the quality of VR content. Implications on LA design for supporting maker education are discussed. Future work is planned to include implementation and evaluation of the dashboard in classrooms.
• Wang, Z., Tzi Dong Ng, J., Liu, R., & Hu, X. (2022). Learning Analytics Enabled Virtual Reality Content Creation Platform: System Design and Preliminary Evaluation. In 22nd International Conference on Advanced Learning Technologies, ICALT 2022.
  More info

  Due to the popularity of virtual reality (VR) in education settings and the rise of maker education, this paper presents LAVR, a platform for VR content creation with learning analytics functions. We design the platform where students can easily create VR stories through a web interface. A learning analytics dashboard is implemented to provide students with feedback on their progress and the quality of the textual content in their VR stories. The platform also offers learning management features for helping teachers set up classrooms with assignments. While the platform will be employed in a forthcoming general education course, we have conducted a preliminary usability evaluation with 12 students and one teacher, and gathered feedback for further refinements before its official launch. The platform will contribute to integrating learning analytics with maker activities.
• Zheng, Y., Que, Y., Hu, X., & Hsiao, J. H. (2022). Predicting Reading Performance based on Eye Movement Analysis with Hidden Markov Models. In 22nd International Conference on Advanced Learning Technologies, ICALT 2022.
  More info

  Reading is an essential medium for learning, but it is challenging to measure learners' cognitive processes during reading. Eye-tracking, as an approach in multimodal learning analytics (MmLA), can provide fine-grained data that reflect cognitive processes during reading. In this study, we investigated whether eye movements could predict passage reading performance in addition to language proficiency and cognitive abilities. In particular, we assessed learners' eye movement pattern and consistency through a novel method, Eye Movement analysis with Hidden Markov Models (EMHMM), in addition to traditional eye movement measures. We found that longer saccade length predicted faster reading speed Also, higher English proficiency predicted faster reading speed through the mediation of longer saccade length. In contrast, reading comprehension accuracy was best predicted by a more consistent eye fixation at the beginning of reading engagement, which may result from a better developed visual routine due to higher reading expertise. These findings have important implications for ways to assess and facilitate learners' reading through eye movement measures and to examine factors influencing reading performance. The methods adopted could further the development of MmLA and serve as an empirical example of understanding learners' cognitive processes through collecting and modeling critical learner-centered metrics in novel modalities.
• Cheong, C. W., Guan, X., & Hu, X. (2021). Using Augmented Reality for Biology Learning in High School: A Quasi-Experiment Study. In 15th International Conference of the Learning Sciences, ICLS 2021.
  More info

  A quasi-experiment was conducted to explore high school students’ perception of Augmented Reality (AR) in biology learning and examine the impact of AR utilization on their academic emotion. The findings indicate students’ general willingness to use AR applications and confirmed the capacity of AR in stimulating positive emotions towards the subject. This suggested that AR can be leveraged to sustain students’ positive academic emotion and engagement in science education.
• Cobb, P. J., Woo, E. M., Pan, N. F., Lou, V. W., Hu, X., Cheng, M., & Xiao, J. (2021). Sharing the Past: The Library as Digital Co-Design Space for Intergenerational Heritage Preservation. In 21st ACM/IEEE Joint Conference on Digital Libraries, JCDL 2021, 2021-.
  More info

  This poster presents the "Intergenerational Participatory Co-design Project, "an interdisciplinary initiative at the University of Hong Kong for facilitating collaboration among different age groups to design digital historic preservation. This project reimagines the global challenge of aging as an opportunity to enhance cultural heritage when older and younger members of society share their unique knowledge and perspectives. Over the course of the 2019-2020 academic year, four mixed-age groups co-designed a variety of innovative digital products to support the preservation and appreciation of Hong Kong's historic culture. The guiding principle of the project was to engage the participants as co-creators of both their own learning outcomes and learning processes. The participants also had opportunities to develop skills with new technologies for documenting, preserving, and presenting cultural heritage. The University of Hong Kong Libraries served as the central space (both physically and virtually) for facilitating these activities, in partnership with the University's Sau Po Centre on Ageing, the Common Core program, and the Faculty of Education. This project can serve as a model for how libraries can support local communities to digitally embrace an aging society for enhancing cultural heritage.
• Hu, X., Ng, J. T., & Lei, C. U. (2021). Evaluation of a lightweight learning analytics tool in moodle and edX: Preliminary results. In 21st IEEE International Conference on Advanced Learning Technologies, ICALT 2021.
  More info

  Learning analytics (LA) mines, analyzes and visualizes the data of students' learning behaviours on learning platforms such as Learning Management Systems (LMS), but few LA tools built are adaptable to multiple platforms or for general education courses. This study sets out to evaluate a lightweight LA tool implemented on Moodle and Open edX for monitoring students' learning progress. Survey data were collected from 156 students, supplemented by interview responses from 25 students and three instructors. Preliminary results show that a considerable portion of surveyed students used the LA tool and they held positive opinions on its efficacy in monitoring self-progress and the effectiveness of its visualizations for information delivery. Nonetheless, learners who did not use the LA tool raised concerns about it relying only on their online behaviours without considering their offline learning. Coupled with instructors' evaluation results, discussion and implications are presented.
• Hu, X., Ng, J. T., & Liu, R. (2021). Development and Evaluation of a Digital Museum of a National Intangible Cultural Heritage from China. In iConference.
  More info

  Intangible cultural heritage (ICH) such as traditional craftsmanship lacks a physical form and often originates from minority groups with little documentation. Digital technologies can be leveraged for documenting and archiving these assets of humanity. In particular, digital museums are established for promoting public understanding and appreciation of cultural heritage. Despite the richness of ICH in China, the development of digital museums of ICH is still in an early stage and mostly from government endeavours. As part of an inter-disciplinary collaborative project involving academic researchers, information professionals, and a private not-for-profit museum, this paper described the development of Gifts from Lanmama, a digital museum of Miao embroidery as a unique ICH from Guizhou ethnic minorities in China. This paper also reported a preliminary evaluation of the digital museum with 78 users, in terms of its usability and affordance for learning about cultural heritage. Results revealed the strengths of the digital museum in terms of the rigor of metadata and its impact on improving users’ understanding and appreciation of Miao embroidery. Some issues and challenges were also identified, such as the lack of channels for user-system communication. These evaluation results offer insights for further improving the digital museum and other end-user oriented digital presentations of similar ICH.
• Hu, X., Ng, J., & Jiang, X. (2021). Young Students’ Experience of Analytics-supported CSCL and the Influence of Parental Attitudes. In 14th International Conference on Computer-Supported Collaborative Learning, CSCL 2021, 2021-.
  More info

  Being recognized as a research priority in the computer-supported collaborative learning (CSCL) community, learning analytics (LA) help harvest and make sense of empirical evidence of students’ collaborative learning. Given that prior investigations on LA-supported CSCL largely focused on university classrooms, implementation of LA for facilitating wiki-based CSCL environments in primary education is rarely explored. The focus on school-aged children also brings about the consideration of their parents’ attitudes. This study aims to examine primary school students’ experience of LA-supported CSCL and how their self-perceived parents’ attitudes towards their use of digital technology influence their learning experience. Survey, wiki logs and interview data were collected from 46 students and three teachers involved in LA-supported wiki-based group inquiry projects. Results show that after receiving LA support, students’ perceptions of inquiry-based learning and those of wiki-supported learning improved while they were less positive towards collaborative learning. In the meantime, it is found that, from students’ perspective, their parents’ awareness of the purpose of their mobile usage was negatively correlated with students’ participation on wiki. Discussion and implications were drawn.
• Li, F., Wang, Z., Tzi Dong Ng, J., & Hu, X. (2021). Studying with learners' own music: Preliminary findings on concentration and task load. In 11th International Conference on Learning Analytics and Knowledge: The Impact we Make: The Contributions of Learning Analytics to Learning, LAK 2021.
  More info

  Through profiling learners' music usage in everyday learning settings and depicting their learning experience when studying with a music app powered by a large-scale and real-world music library, this study revealed preliminary observations on how background music impacts learning under varying task load, and manifested intriguing patterns of learners' music usage and music preferences in various task load conditions. Specifically, we piloted a three-day field experiment in students' everyday learning environment. During the experiment, participants performed learning tasks with music in the background and completed a set of online surveys before and after each learning session. Our results suggested that learners' self-selected, real-life background music could enhance their learning effectiveness, while the beneficial effect of background music was more apparent when the learning task was less mentally or temporally demanding. Towards a closer look at the characteristics of preferable music pieces under various task load conditions, our findings showed that music preferred by participants under high versus low temporal demand differs in a number of characteristics, including speechiness, acousticness, danceability, and energy. This study further reveals the effects of background music on learning under varying task load levels and provides implications for context-aware background music selection when designing musically enriched learning environments.
• Li, F., Xiao, Z., Ng, J. T., & Hu, X. (2021). Exploring Interdisciplinary Data Science Education for Undergraduates: Preliminary Results. In iConference.
  More info

  This paper reports a systematic literature review on undergraduate data science education followed by semi-structured interviews with two frontier data science educators. Through analyzing the hosting departments, design principles, curriculum objectives, and curriculum design of existing programs, our findings reveal that (1) the data science field is inherently interdisciplinary and requires joint collaborations between various departments. Multi-department administration was one of the solutions to offer interdisciplinary training, but some problems have also been identified in its practical implementation; (2) data science education should emphasize hands-on practice and experiential learning opportunities to prepare students for data analysis and problem-solving in real-world contexts; and (3) although the importance of comprehensive coverage of various disciplines in data science curricula is widely acknowledged, how to achieve an effective balance between various disciplines and how to effectively integrate domain knowledge into the curriculum still remain open questions. Findings of this study can provide insights for the design and development of emerging undergraduate data science programs.
• Que, Y., & Hu, X. (2021). Investigate the effects of background music on visual cognitive tasks using multimodal learning analytics. In 21st IEEE International Conference on Advanced Learning Technologies, ICALT 2021.
  More info

  Music is a popular form of entertainment and has become common practice to adjust cognition, affect, and motivation. Regarding the effects of background music on learning tasks, results are inconclusive in the literature. Recent advancement of wearable devices and computing analytics supports automated detection of multimodal physiological signals, such as eye movements, neural responses, and heart rates in a real-time fashion, which can facilitate tracking learners' changes of affect, attention, and cognition while they study with the accompaniment of background music. However, most existing studies focus only on behavioral levels, and few employed signals at physiological levels to investigate the impact of background music on learning. To fill in the research gap, this doctoral project designs two types of visual cognitive tasks, that is, reading comprehension task and art appreciation task. It aims to integrate multimodal data (e.g., eye movements, electroencephalogram (EEG), and peripheral physiological signals) to probe the effect of background music on the tasks. Its findings will extend our knowledge on the interactions among learners' performance, emotion, and engagement at both physiological and behavioral levels in multi-channel learning settings, and contribute to a goal of recommending suitable background music for self-learning.
• Wang, C., Li, Q., & Hu, X. (2021). Minority college students' engagement in learning activities and its relationships with learning outcomes. In 21st IEEE International Conference on Advanced Learning Technologies, ICALT 2021.
  More info

  Minority college students' learning engagement is an important perspective for improving education equity and facilitating learning. Based on responses from a large-scale experience survey among Chinese college students, this study aims to explore minority students' engagement in nine learning activities and its relationship with learning outcomes, as well as their difference from overall student sample. Descriptive statistical analysis, single sample t-test, and multiple linear regression analysis were conducted. The results revealed: (1) Minority students' engagement in learning activities was low, especially in formal learning activities. (2) Interaction with teachers, engagement in arts activities did not predict minority students' learning outcomes, which was different from findings in the overall student sample. (3) Engagement in student organizations did not predict students' learning outcomes in either sample. Based on results, suggestions were discussed for improving students' learning engagement.
• Zhou, Z., Tam, V. W., Lui, K. S., Lam, E. Y., Kong, R., Hu, X., & Law, N. (2021). A new approach for educational data analytics with wearable devices. In 21st IEEE International Conference on Advanced Learning Technologies, ICALT 2021.
  More info

  The rapid development of wearable technologies has dramatically promoted the potential usages of wearable devices in educational data analytics. However, the large amount of input data and the various types of educational output labels also increase the difficulties in selecting the useful information and discovering the implicit relations between different input data. To address this issue, this paper proposed a new two-layer approach for conducting educational data analytics automatically. In this approach, there are three key components: input layer, output layer and recognition model. For the input layer, we adopted the newly proposed optimization algorithm: Adaptive Multi-Population Optimization (AMPO) to select the most related input features and suitable model structures. For the output layer, we inserted domain-specific constraints during the searching for all combinations of different output labels to discover a meaningful output strategy with a relatively higher accuracy. Based on the input elements and output strategy provided by the input layer and the output layer, the recognition model will produce the corresponding recognition accuracy. With these three components, our proposed method can find out some connotative information to provide guidance for conducting educational data analytics and drawing meaningful conclusions.
• Hu, X., Ng, J. T., Yang, C., & Chu, S. K. (2020). Personalized book recommendation to young readers: Two online prototypes and a preliminary user evaluation. In 2020 ACM/IEEE-CS Joint Conference on Digital Libraries, JCDL 2020.
  More info

  Online learning platforms that aim to improve reading interests and proficiency of young readers, particularly students in elementary schools, rarely have automated personalized recommendation services. This study attempts to bridge this gap by developing and evaluating two book recommenders that are integrated into an online learning platform for young readers. A preliminary user experiment was conducted to measure the effectiveness and usability of the recommender prototypes. Results of think-aloud usability testing, post-test questionnaires, and a semi-structured interview verified the feasibility of adding these book recommenders to improve personalization of the online learning platform. Further improvements of the recommenders were also suggested. Th e user evaluation framework provides a reference for future studies on personalized learning material recommendation.
• Kong, R., Hu, X., & Yuen, A. H. (2020). Understanding academic engagement and context through multimodal data. In 53rd Annual Hawaii International Conference on System Sciences, HICSS 2020, 2020-.
  More info

  The Internet has penetrated the life of adolescents and become a new space for learning, socializing and entertainment. Physical exercise and sleep remain crucial for the development of adolescents. However, the influence of these critical contextual factors on learning and life is insufficiently explored, partially due to the difficulty of measuring these factors. To bridge the gap, a semi-automated Day Reconstruction Method was proposed which leverages 24-hour multimodal data collected by smart wristband (Fitbit Versa), paired mobile phone (Red Mi 6), and time management application (RescueTime). A pilot study was conducted to verify the feasibility of this proposed method and is reported in this paper. As academic engagement among adolescents is widely concerned by stakeholders, this pilot study also explores the relationship between the aforementioned contextual factors and academic engagement. With some interesting patterns, revealed, this study contributes to furthering our understanding of how context shapes adolescents' academic engagement using a more objective and nonintrusive method.
• Li, F., Hu, X., & Que, Y. (2020). Learning with background music: A field experiment. In 10th International Conference on Learning Analytics and Knowledge: Shaping the Future of the Field, LAK 2020.
  More info

  Empirical evidence of how background music benefits or hinders learning becomes the crux of optimizing music recommendation in educational settings. This study aims to further probe the underlying mechanism through an experiment in naturalistic setting. 30 participants were recruited to join a field experiment which was conducted in their own study places for one week. During the experiment, participants were asked to conduct learning sessions with music in the background and collect music tracks they deemed suitable for learning using a novel mobilebased music discovery application. A set of participant-related, context-related, and music-related data were collected via a preexperiment questionnaire, surveys popped up in the music app, and the logging system of the music app. Preliminary results reveal correlations between certain music characteristics and learners' task engagement and perceived task performance. This study is expected to provide evidence for understanding cognitive and emotional dimensions of background music during learning, as well as implications for the role of personalization in the selection of background music for facilitating learning.
• Liu, M., Zangerle, E., Hu, X., Melchiorre, A., & Schedl, M. (2020). PANDEMICS, MUSIC, AND COLLECTIVE SENTIMENT: EVIDENCE FROM THE OUTBREAK OF COVID-19. In 21st International Society for Music Information Retrieval Conference, ISMIR 2020.
  More info

  The COVID-19 pandemic causes a massive global health crisis and produces substantial economic and social distress, which in turn may cause stress and anxiety among people. Real-world events play a key role in shaping collective sentiment in a society. As people listen to music daily everywhere in the world, the sentiment of music being listened to can reflect the mood of the listeners and serve as a measure of collective sentiment. However, the exact relationship between real-world events and the sentiment of music being listened to is not clear. Driven by this research gap, we use the unexpected outbreak of COVID-19 as a natural experiment to explore how users’ sentiment of music being listened to evolves before and during the outbreak of the pandemic. We employ causal inference approaches on an extended version of the LFM-1b dataset of listening events shared on Last.fm, to examine the impact of the pandemic on the sentiment of music listened to by users in different countries. We find that, after the first COVID-19 case in a country was confirmed, the sentiment of artists users listened to becomes more negative. This negative effect is pronounced for males while females’ music emotion is less influenced by the outbreak of the COVID-19 pandemic. We further find a negative association between the number of new weekly COVID-19 cases and users’ music sentiment. Our results provide empirical evidence that public sentiment can be monitored based on collective music listening behaviors, which can contribute to research in related disciplines.
• Liu, R., & Hu, X. (2020). A multimodal music recommendation system with listeners' personality and physiological signals. In 2020 ACM/IEEE-CS Joint Conference on Digital Libraries, JCDL 2020.
  More info

  This preliminary study explored multiple information sources for music recommendation system (MRS), including users' personality traits measured by the Ten-Item Personality Inventory (TIPI) and physiological signals recorded by a wearable wristband. A dataset of 23 participants and 628 song listening records were obtained from a user experiment, with matched personality, physiological signals as well as music acoustic features. Based on the dataset, a machine learning experiment with four regression algorithms was conducted to compare recommendation performances across different combinations of feature sets. Results show that personality features contributed significantly to the improvement of recommender accuracy, while physiological features contributed less. Analysis of top features in the best performing model revealed the importance of some physiological features. Future studies are called for to further investigate multimodal MRS through exploiting user properties and context data.
• Ng, J. T., Ng, W., Hu, X., & Jung, T. P. (2020). Evaluation of low-end virtual reality content of cultural heritage: A preliminary study with eye movement. In 2020 ACM/IEEE-CS Joint Conference on Digital Libraries, JCDL 2020.
  More info

  The affordances of virtual reality (VR) have made it widely adopted for presenting the content of cultural heritage in digital libraries. In recent years, non-specialists including university students were involved in creating low-end VR content using low-cost equipment and software. Among various design options, the question as to which ones could be more effective in presenting cultural heritage remains. This study aims to evaluate and compare the effectiveness of user-created VR content of cultural heritage with different designs, through collecting and analyzing self-report and eye movement data from end users. Results show that the presence of text annotations in the VR content helped users understand the cultural heritage being presented, whereas users' visual attention was largely attracted to the text annotations and additional images when the VR content contained such visual information. This preliminary study also explores the feasibility of using the eye-tracking method to analyze user interactions with VR content of cultural heritage. The results provide empirical evidence on the effects of different designs of user-created VR content on end users' understanding of cultural heritage.
• Que, Y., Zheng, Y., Hsiao, J. H., & Hu, X. (2020). Exploring the effect of personalized background music on reading comprehension. In 2020 ACM/IEEE-CS Joint Conference on Digital Libraries, JCDL 2020.
  More info

  It is a common phenomenon that many students study with background music, but the influence of background music on learning is still an open question, with inconclusive findings in the literature. Inspired by the research gap, we conducted a controlled user experiment on reading with 100 students from a comprehensive university. The participants were tasked to read nine academic passages. In the meantime, those who were randomly allocated to the experiment group listened to their self-provided music in the background during the reading task, while those in the control group did not have background music during reading. During the experiment, participants' reading logs, self-reported meta-cognition and emotion status were recorded. This paper reports the results of comparing measures on reading performance, meta-cognition and emotion changes between the two groups. In addition, the relationships between participants' personal traits and their preferred background music types were investigated. Findings indicated that learning with background music of one's own choice could be beneficial for maintaining positive emotion, with no cost on reading performance. Through providing empirical evidence on the effect of background music on reading, this study contributes to furthering our understanding on human behaviors in multichannel learning settings and rendering design implications for personalized recommendations in online music services and music digital libraries for facilitating reading and self-learning.
• Zhou, Z. X., Tam, V., Lui, K. S., Lam, E. Y., Hu, X., Yuen, A., & Law, N. (2020). A sophisticated platform for learning analytics with wearable devices. In 20th IEEE International Conference on Advanced Learning Technologies, ICALT 2020.
  More info

  With the rapid development in wearable technology, wearable devices integrating with various sensors have been broadly applied in different areas. Yet there is seldom any previous study which focuses on applying wearable devices and deep learning in learning analytics. This paper considers a sophisticated real-time learning analytics platform for analyzing students' learning states and learning activities with wearable devices and deep learning. During the experimental period of this platform, students will receive instant notifications from an intelligent mobile application when their heart rate are out of their normal range so that the actual learning activities conducted by students can be collected to train deep learning models for recognizing their learning activities. At the same time, students can enjoy the sleeping monitoring and the exercise monitoring functionalities provided by the smart watches in this platform. The results of the interviews conducted after the experiment for this platform demonstrate that 89% of students think that this platform is useful for their daily lives and 65% of students report that this platform brings positive effects on their learning in different aspects. More importantly, this work sheds lights on the possibility of applying wearable devices in learning analytics to improve the learning effectivenesses and life qualities of students.
• Hu, X., Li, F., & Kong, R. (2019). Can background music facilitate learning? Preliminary results on reading comprehension. In 9th International Conference on Learning Analytics and Knowledge, LAK 2019.
  More info

  It is a common phenomenon for students to listen to background music while studying. However, there are mixed and inconclusive Kindings in the literature, leaving it unclear whether and in which circumstances background music can facilitate or hinder learning. This paper reports a study investigating the effects of Kive different types of background audio (four types of music and one environmental sound) on reading comprehension. An experiment was conducted with 33 graduate students, where a series of cognitive, metacognitive, affective variables and physiological signals were collected and analyzed. Preliminary results show that there were differences on these variables across different music types. This study contributes to the understanding and optimizing of background music for facilitating learning.
• Hu, X., Que, Y., Kando, N., & Lian, W. (2019). Analyzing user interactions with music information retrieval system: An eye-tracking approach. In 20th International Society for Music Information Retrieval Conference, ISMIR 2019.
  More info

  There has been little research considering eye movement as a measure when assessing user interactions with music information retrieval (MIR) systems, whereas many studies have adopted conventional user-centered measures such as user effectiveness and user perception. To bridge this research gap, this study investigates users' eye movement patterns and measures with two music retrieval tasks and two interface presentation modes. A user experiment was conducted with 16 participants whose eye movement and mouse click behaviors were recorded through professional eye trackers. Through analyzing visual patterns of eye gazes and movements as well as various metrics in prominent Areas of Interest (AOI), it is found that users' eye movement behaviors were related to task type. Besides, the results also disclosed that some eye movement metrics were related to both user effectiveness and user perception, and influenced by user characteristics. It is also found that some eye movement and user effectiveness metrics can be used to predict user perception. This study allows researchers to gain a deeper insight into user interactions with MIR systems from the perspective of eye movement measure.
• Kostiuk, B., Costa, Y. M., Britto, A. S., Hu, X., & Silla, C. N. (2019). Multi-label emotion classification in music videos using ensembles of audio and video features. In 31st IEEE International Conference on Tools with Artificial Intelligence, ICTAI 2019, 2019-.
  More info

  Video as well as music are potent means to convey emotions. However, despite their importance in several applications, few works deal with the issue of emotion classification in videos. The main reason is possibly the lack of available databases. In this work we extend the CAL500 database by including music videos, since the CAL500 was originally proposed as an audio-only database. The main rationale here is that the music videos must be official as they were developed to convey the same emotion as the song. After adapting the database, we have extracted audio and video features to perform our computational experiments. Our main result is that there is a complementarity between the audio and video features as the best result was achieved using their combination.
• Ng, J., Lei, C. U., Lau, E., Lui, K. S., Lam, K. H., Kwok, T. T., Hu, X., Warning, P., & Tam, V. (2019). Applying Instructional Design in Engineering Education and Industrial Training: An Integrative Review. In 2019 IEEE International Conference on Engineering, Technology and Education, TALE 2019.
  More info

  Various pedagogies have been proposed in the engineering education literature. However, simply applying pedagogical approaches might be ineffective in helping learners acquire and exhibit skills necessary for future industrial endeavours, if learning activities and learning objects are not structurally designed. In this paper, we present an integrative and interpretive review of studies to show how instructional design principles listed in the First Principles of Instruction (FPI) framework are in alignment with teaching and learning of and with cognitive and psychomotor skills, particularly in industrial training and engineering education. The applications of FPI and its impacts on learning were discussed, whereas congruence between the i) FPI and the Bloom's Taxonomy, and ii) FPI and psychomotor instruction, were identified. This study aims to offer insights for redesigning teaching and training in engineering education with an effective instructional design, ultimately enhancing the pedagogical values of its learning objects and activities.
• Qiao, C., & Hu, X. (2019). Measuring knowledge gaps in student responses by mining networked representations of texts. In 9th International Conference on Learning Analytics and Knowledge, LAK 2019.
  More info

  Gaps between knowledge sources are interesting to various stakeholders: they might indicate potential misconceptions awaiting correction, complex or novel knowledge that requires careful delivery or studying. Motivated by these underlying values, this study explores the knowledge gap phenomenon in the context of student textual responses. In the method proposed in this study, discourses are ]irst mapped into structured knowledge spaces where gaps between correct/incorrect responses and assessed knowledge are measured by network-based metrics. Empirical results demonstrate the effectiveness of the proposed method in measuring gaps in student responses. The networked representation of texts proposed in this study is novel in quantitatively framing gaps of knowledge. It also offers a set of validated metrics for analyzing student responses in research and practice.
• Zhou, Z. X., Tam, V., Lui, K. S., Lam, E. Y., Yuen, A., Hu, X., & Law, N. (2019). Applying deep learning and wearable devices for educational data analytics. In 31st IEEE International Conference on Tools with Artificial Intelligence, ICTAI 2019, 2019-.
  More info

  With the popularity of wearable devices, smart watches containing various sensors have been widely adopted for many healthcare applications. Yet there is rarely any research study on the possible uses of smart watches for learning analytics, particularly for analyzing students' learning activities through the physiological and/or movement data collected on their smart watches. This paper considers a pioneering and sophisticated learning analytics platform using fine-tuned deep learning models to predict students' learning activities based on the real-time data, including their heart rates, calories, three-axis accelerometer and gyroscope data, captured on wearable devices and then uploaded onto a cloud server for thorough analyses. To validate on the actual activities conducted by each student, an intelligent mobile application is developed to push instant notifications for students to report their own activities whenever the change of heart rates are deviated significantly from their normal values. Based on students' heart rates and calories, a long-short term memory (LSTM) model is built to classify students' learning states as active or not with an impressive prediction accuracy of 95% whereas another hybrid model combining both the LSTM and convolutional neural networks attains the highest prediction accuracy of 74% to predict students' specific learning activities as based on their physiological and movement data. The prototype implementation clearly demonstrates the feasibility of the proposed framework for learning analytics. More importantly, this work shed lights on various directions including the integration of noise filters to preprocess the collected data for further investigation.
• Chai, Y., Lei, C. U., Hu, X., & Kwok, Y. K. (2018). WPSS: Dropout Prediction for MOOCs using course progress normalization and subset selection. In 5th Annual ACM Conference on Learning at Scale, L at S 2018.
  More info

  There are existing multi-MOOC level dropout prediction research in which many MOOCs' data are involved. This generated good results, but there are two potential problems. On one hand, it is inappropriate to use which week students are in to select training data because courses are with different durations. On the other hand, using all other existing data can be computationally expensive and inapplicable in practice. To solve these problems, we propose a model called WPSS (WPercent and Subset Selection) which combines the course progress normalization parameter wpercent and subset selection. 10 MOOCs offered by The University of Hong Kong are involved and experiments are in the multi-MOOC level. The best performance of WPSS is obtained in neural network when 50% of training data is selected (average AUC of 0.9334). Average AUC is 0.8833 for traditional model without wpercent and subset selection in the same dataset.
• Hu, X., Li, F., & Ng, J. T. (2018). On the relationships between music-induced emotion and physiological signals. In 19th International Society for Music Information Retrieval Conference, ISMIR 2018.
  More info

  Emotion-aware music information retrieval (MIR) has been difficult due to the subjectivity and temporality of emotion responses to music. Physiological signals are regarded as related to emotion and thus could potentially be exploited in emotion-aware music discovery. This study explored the possibility of using physiological signals to detect users’ emotion responses to music, with consideration of individual characteristics (personality, music preferences, etc.). A user experiment was conducted with 23 participants who searched for music in a novel MIR system. Users’ listening behaviors and self-reported emotion responses to a total of 628 music pieces were collected. During music listening, a series of peripheral physiological signals (e.g., heart rate, skin conductance) were recorded from participants unobtrusively using a research-grade wearable wristband. A set of features in the time- and frequency- domains were extracted from the physiological signals and analyzed using statistical and machine learning methods. Results reveal 1) significant differences in some physiological features between positive and negative arousal and mood categories, and 2) effective classification of emotion responses based on physiological signals for some individuals. The findings can contribute to further improvement of emotion-aware intelligent MIR systems exploiting physiological signals as an objective and personalized input.
• Hu, X., Tam, I. K., Liu, M., & Downie, J. S. (2018). Music artist similarity: An exploratory study on a large-scale dataset of online streaming services. In iConference.
  More info

  In supporting music search, online music streaming services often suggest artists who are deemed as similar to those listened to or liked by users. However, there has been an ongoing debate on what constitutes artist similarity. Approaching this problem from an empirical perspective, this study collected a large-scale dataset of similar artists recommended in four well-known online music steaming services, namely Spotify, Last.fm, the Echo Nest, and KKBOX, on which an exploratory quantitative analysis was conducted. Preliminary results reveal that similar artists in these services were related to the genre and popularity of the artists. The findings shed light on how the concept of artist similarity is manifested in widely adopted real-world applications, which will in turn help enhance our understanding of music similarity and recommendation.
• Qiao, C., & Hu, X. (2018). Discovering student behavior patterns from event logs: Preliminary results on a novel probabilistic latent variable model. In 18th IEEE International Conference on Advanced Learning Technologies, ICALT 2018.
  More info

  Digital platforms enable the observation of learning behaviors through fine-grained log traces, offering more detailed clues for analysis. In addition to previous descriptive and predictive log analysis, this study aims to simultaneously model learner activities, event time spans, and interaction levels using the proposed Hidden Behavior Traits Model (HBTM). We evaluated model performance and explored their capability of clustering learners on a public dataset, and tried to interpret the machine recognized latent behavior patterns. Quantitative and qualitative results demonstrated the promising value of HBTM. Results of this study can contribute to the literature of online learner modeling and learning service planning.
• Hu, X., Cheong, C. W., Ding, W., & Woo, M. (2017). A systematic review of studies on predicting student learning outcomes using learning analytics. In 7th International Conference on Learning Analytics and Knowledge, LAK 2017.
  More info

  Predicting student learning outcomes is one of the prominent themes in Learning Analytics research. These studies varied to a significant extent in terms of the techniques being used, the contexts in which they were situated, and the consequent effectiveness of the prediction. This paper presented the preliminary results of a systematic review of studies in predictive learning analytics. With the goal to find out what methodologies work for what circumstances, this study will be able to facilitate future research in this area, contributing to relevant system developments that are of pedagogic values.
• Hu, X., Choi, K., Hao, Y., Cunningham, S. J., Lee, J. H., Laplante, A., Bainbridge, D., & Downie, J. S. (2017). Exploring the music library association mailing list: A text mining approach. In 18th International Society for Music Information Retrieval Conference, ISMIR 2017.
  More info

  Music librarians and people pursuing music librarianship have exchanged emails via the Music Library Association Mailing List (MLA-L) for decades. The list archive is an invaluable resource to discover new insights on music information retrieval from the perspective of the music librarian community. This study analyzes a corpus of 53,648 emails posted on MLA-L from 2000 to 2016 by using text mining and quantitative analysis methods. In addition to descriptive analysis, main topics of discussions and their trends over the years are identified through topic modeling. We also compare messages that stimulated discussions to those that did not. Inspection of semantic topics reveals insights complementary to previous topic analyses of other Music Information Retrieval (MIR) related resources.
• Hu, X., Hou, X., Lei, C. U., Yang, C., & Ng, J. (2017). An outcome-based dashboard for Moodle and Open edX. In 7th International Conference on Learning Analytics and Knowledge, LAK 2017.
  More info

  This poster presents a cross-platform learning analytics dashboard on Moodle and Open edX for monitoring outcome-based learning progress. The dashboard visualizes students' interactions with the platforms in near real-time, aiming to help teachers and students monitor students' learning progress. The dashboard has been used in four large-size general education courses in a comprehensive university in Hong Kong, undergoing evaluation and improvement.
• Hu, X., Yang, C., Qiao, C., Lu, X., & Chu, S. K. (2017). New features in wikiglass, a learning analytic tool for visualizing collaborative work on wikis. In 7th International Conference on Learning Analytics and Knowledge, LAK 2017.
  More info

  Wikiglass is a learning analytic tool for visualizing collaborative work on Wikis built by groups of secondary or primary school students. This poster presents new features of Wikiglass developed recently based on requests from teachers, including flexible selection of date range, revision network, and thinking order detection. Currently the new features are used and evaluated in two secondary schools in Hong Kong.
• Liu, M., Hu, X., & Schedl, M. (2017). Artist preferences and cultural, socio-economic distances across countries: A big data perspective. In 18th International Society for Music Information Retrieval Conference, ISMIR 2017.
  More info

  Users in different countries may have different music preferences, possibly due to geographical, economic, linguistic, and cultural factors. Revealing the relationship between music preference and cultural socio-economic differences across countries is of great importance for music information retrieval in a cross-country or cross-cultural context. Existing works are usually based on small samples in one or several countries or take only one or two socio-economic aspects into account. To bridge the gap, this study makes use of a large-scale music listening dataset, LFM-1b with more than one billion music listening logs, to explore possible associations between a variety of cultural and socio-economic measurements and artist preferences in 20 countries. From a big data perspective, the results reveal: 1) there is a highly uneven distribution of preferred artists across countries; 2) the linguistic differences among these countries are positively associated with the distances in artist preferences; 3) country differences in three of the six cultural dimensions considered in this study have positive influences on the difference of artist preferences among the countries; and 4) geographical and economic distances among the countries have no significant relationship with their artist preferences.
• Hu, X., Choi, K., Lee, J. H., Laplante, A., Hao, Y., Cunningham, S. J., & Downie, J. S. (2016). WIMIR: An informetric study on women authors in ISMIR. In 17th International Society for Music Information Retrieval Conference, ISMIR 2016.
  More info

  The Music Information Retrieval (MIR) community is becoming increasingly aware of a gender imbalance evident in ISMIR participation and publication. This paper reports upon a comprehensive informetric study of the publication, authorship and citation characteristics of female researchers in the context of the ISMIR conferences. All 1,610 papers in the ISMIR proceedings written by 1,910 unique authors from 2000 to 2015 were collected and analyzed. Only 14.1% of all papers were led by female researchers. Temporal analysis shows that the percentage of lead female authors has not improved over the years, but more papers have appeared with female coauthors in very recent years. Topics and citation numbers are also analyzed and compared between female and male authors to identify research emphasis and to measure impact. The results show that the most prolific authors of both genders published similar numbers of ISMIR papers and the citation counts of lead authors in both genders had no significant difference. We also analyzed the collaboration patterns to discover whether gender is related to the number of collaborators. Implications of these findings are discussed and suggestions are proposed on how to continue encouraging and supporting female participation in the MIR field.
• Hu, X., Ip, J., Sadaful, K., Lui, G., & Chu, S. (2016). Wikiglass: A learning analytic tool for visualizing collaborative wikis of secondary school students. In 6th International Conference on Learning Analytics and Knowledge, LAK 2016, 25-29-.
  More info

  This demo presents Wikiglass, a learning analytic tool for visualizing the statistics and timelines of collaborative Wikis built by secondary school students during their group project in inquiry-based learning. The tool adopts a modular structure for the flexibility of reuse with different data sources. The client side is built with the Model-View-Controller framework and the AngularJS library whereas the server side manages the database and data sources. The tool is currently used by secondary teachers in Hong Kong and is undergoing evaluation and improvement.
• Hu, X., Ng, T. D., Tian, L., & Lei, C. U. (2016). Automating assessment of collaborative writing quality in multiple stages: The case of wiki. In 6th International Conference on Learning Analytics and Knowledge, LAK 2016, 25-29-.
  More info

  This study attempts to investigate to what extent indicators of academic writing and cognitive thinking can help measure the writing quality of group collaborative writings on Wikis. Particularly, comparisons were made on Wiki content in different stages of the projects. Preliminary results from a multiple linear regression analysis reveal that linguistic indicators such as engagement markers and self-mention were significant predictors in earlier stages to the projects, whereas verbs indicating cognitive thinking in the evaluation level were significant in later project stages.
• Hu, X., Zhang, Y., Chu, S. K., & Ke, X. (2016). Towards personalizing An E-quiz bank for primary school students: An exploration with association rule mining and clustering. In 6th International Conference on Learning Analytics and Knowledge, LAK 2016, 25-29-.
  More info

  Given the importance of reading proficiency and habits for young students, an online e-quiz bank, Reading Battle, was launched in 2014 to facilitate reading improvement for primary-school students. With more than ten thousand questions in both English and Chinese, the system has attracted nearly five thousand learners who have made about half a million question answering records. In an effort towards delivering personalized learning experience to the learners, this study aims to discover potentially useful knowledge from learners' reading and question answering records in the Reading Battle system, by applying association rule mining and clustering analysis. The results show that learners could be grouped into three clusters based on their self-reported reading habits. The rules mined from different learner clusters can be used to develop personalized recommendations to the learners. Implications of the results on evaluating and further improving the Reading Battle system are also discussed.
• Lee, J. H., Hu, X., Choi, K., & Downie, J. S. (2015). Mirex grand challenge 2014 user experience: Qualitative analysis of user feedback. In 16th International Society for Music Information Retrieval Conference, ISMIR 2015.
  More info

  Evaluation has always been fundamental to the Music Information Retrieval (MIR) community, as evidenced by the popularity of the Music Information Retrieval Evaluation eXchange (MIREX). However, prior MIREX tasks have primarily focused on testing specialized MIR algorithms that sit on the back end of systems. Not until the Grand Challenge 2014 User Experience (GC14UX) task had the users’ overall interaction and experience with complete systems been formally evaluated. Three systems were evaluated based on five criteria. This paper reports the results of GC14UX, with a special focus on the qualitative analysis of 99 free text responses collected from evaluators. The analysis revealed additional user opinions, not fully captured by score ratings on the given criteria, and demonstrated the challenge of evaluating a variety of systems with different user goals. We conclude with a discussion on the implications of findings and recommendations for future UX evaluation tasks, including adding new criteria: Aesthetics, Performance, and Utility.
• Bainbridge, D., Hu, X., & Downie, J. S. (2014). A musical progression with greenstone: How music content analysis and linked data is helping redefine the boundaries to a music digital library. In 1st International Workshop on Digital Libraries for Musicology, DLfM 2014, 12-.
  More info

  Despite the recasting of the web's technical capabilities through Web 2.0, conventional digital library software architectures-from which many of our leading Music Digital Libraries (MDLs) are formed-result in digital resources that are, surprisingly, disconnected from other online sources of information, and embody a "read-only" mindset. Leveraging from Music Information Retrieval (MIR) techniques and Linked Open Data (LOD), in this paper we demonstrate a new form of music digital library that encompasses management, discovery, delivery, and analysis of the musical content it contains. Utilizing open source tools such as Greenstone, audioDB, Meandre, and Apache Jena we present a series of transformations to a musical digital library sourced from audio files that steadily increases the level of support provided to the user for musicological study. While the seed for this work was motivated by better supporting musicologists in a digital library, the developed software architecture alters the boundaries to what is conventionally thought of as a digital library- and in doing so challenges core assumptions made in mainstream digital library software design.
• Deng, S., & Hu, X. (2014). Creating a Knowledge Map for the research lifecycle. In KMIR 2014 - 1st Workshop on Knowledge Maps and Information Retrieval, co-located with International Conference on Digital Libraries, DL 2014 - ACM/IEEE Joint Conference on Digital Libraries, JCDL 2014 and International Conference on Theory and Practice of Digital Libraries, TPDL 2014, 1311.
  More info

  In this study, a Knowledge Map (KM) was created based on the Research Lifecycle at the University of Central Florida to provide campus-wide services and resources to researchers. The KM aims to meet the needs of researchers and delivers guided searching and assistance in all aspects of research, including literature review, citation management, research data management, grant management, research work publication and dissemination. It elaborates the research processes and their associated services as presented in the Research Lifecycle, and links these points to various campus resources including those provided by the University Libraries, the Office of Research and Commercialization, the Institute for Simulation and Training and the Faculty Center for Teaching and Learning. It gives unified support to the researchers during their entire research lifecycle and it will keep evolving and developing.
• Hu, X., & Yang, Y. H. (2014). Cross-cultural mood regression for music digital libraries. In 2014 14th IEEE/ACM Joint Conference on Digital Libraries, JCDL 2014.
  More info

  Mood is a popular access point in music digital libraries and online music repositories, and is often represented as numerical values in a small number of emotion-related dimensions (e.g., valence and arousal). As music mood is recognized as culturally dependent, this study investigates whether regression models built with music data in one culture can be applied to music in another culture. Results indicate that cross-cultural predictions of both valence and arousal values are feasible.
• Hu, X., & Yang, Y. H. (2014). La study on cross-cultural and cross-dataset generalizability of music mood regression models. In 40th International Computer Music Conference, ICMC 2014, Joint with the 11th Sound and Music Computing Conference, SMC 2014 - Music Technology Meets Philosophy: From Digital Echos to Virtual Ethos.
  More info

  The goal of music mood regression is to represent the emotional expression of music pieces as numerical values in a low-dimensional mood space and automatically predict those values for unseen music pieces. Existing studies on this topic usually train and test regression models using music datasets sampled from the same culture source, annotated by people with the same cultural background, or otherwise constructed by the same method. In this study, we explore whether and to what extent regression models trained with samples in one dataset can be applied to predicting valence and arousal values of samples in another dataset. Specifically, three datasets that differ in factors such as cultural backgrounds of stimuli (music) and subjects (annotators), stimulus types and annotation methods are evaluated and the results suggested that cross-cultural and cross-dataset predictions of both valence and arousal values could achieve comparable performance to within-dataset predictions. We also discuss how the generalizability of regression models can be affected by dataset characteristics. Findings of this study may provide valuable insights into music mood regression for non-Western and other music where training data are scarce.
• Hu, X., Lee, J. H., & Wong, L. K. (2014). Music information behaviors and system preferences of university students in Hong kong. In 15th International Society for Music Information Retrieval Conference, ISMIR 2014.
  More info

  This paper presents a user study on music information needs and behaviors of university students in Hong Kong. A mix of quantitative and qualitative methods was used. A survey was completed by 101 participants and supplemental interviews were conducted in order to investigate users’ music information related activities. We found that university students in Hong Kong listened to music frequently and mainly for the purposes of entertainment, singing and playing instruments, and stress reduction. This user group often searches for music with multiple methods, but common access points like genre and time period were rarely used. Sharing music with people in their online social networks such as Facebook and Weibo was a common activity. Furthermore, the popularity of smartphones prompted the need for streaming music and mobile music applications. We also examined users’ preferences on music services available in Hong Kong such as YouTube and KKBox, as well as the characteristics liked and disliked by the users. The results not only offer insights into non-Western users’ music behaviors but also for designing online music services for young music listeners in Hong Kong.
• Hu, X., Lee, J. H., Choi, K., & Downie, J. S. (2014). A cross-cultural study of mood in K-pop songs. In 15th International Society for Music Information Retrieval Conference, ISMIR 2014.
  More info

  Prior research suggests that music mood is one of the most important criteria when people look for music—but the perception of mood may be subjective and can be influenced by many factors including the listeners’ cultural background. In recent years, the number of studies of music mood perceptions by various cultural groups and of automated mood classification of music from different cultures has been increasing. However, there has yet to be a well-established testbed for evaluating cross-cultural tasks in Music Information Retrieval (MIR). Moreover, most existing datasets in MIR consist mainly of Western music and the cultural backgrounds of the annotators were mostly not taken into consideration or were limited to one cultural group. In this study, we built a collection of 1,892 K-pop (Korean Pop) songs with mood annotations collected from both Korean and American listeners, based on three different mood models. We analyze the differences and similarities between the mood judgments of the two listener groups, and propose potential MIR tasks that can be evaluated on this dataset.
• Stephen Downie, J., Hu, X., Lee, J. H., Choi, K., Cunningham, S. J., & Hao, Y. (2014). Ten years of MIREX: Reflections, challenges and opportunities. In 15th International Society for Music Information Retrieval Conference, ISMIR 2014.
  More info

  The Music Information Retrieval Evaluation eXchange (MIREX) has been run annually since 2005, with the October 2014 plenary marking its tenth iteration. By 2013, MIREX has evaluated approximately 2000 individual music information retrieval (MIR) algorithms for a wide range of tasks over 37 different test collections. MIREX has involved researchers from over 29 different countries with a median of 109 individual participants per year. This paper summarizes the history of MIREX from its earliest planning meeting in 2001 to the present. It reflects upon the administrative, financial, and technological challenges MIREX has faced and describes how those challenges have been surmounted. We propose new funding models, a distributed evaluation framework, and more holistic user experience evaluation tasks—some evolutionary, some revolutionary—for the continued success of MIREX. We hope that this paper will inspire MIR community members to contribute their ideas so MIREX can have many more successful years to come.
• Zhang, H., & Hu, X. (2014). A quantitative comparison on file folder structures of two groups of information workers. In 2014 14th IEEE/ACM Joint Conference on Digital Libraries, JCDL 2014.
  More info

  This study compares file folder structures on personal computers of two groups of information workers, administrative staff and PhD students. A set of quantitative measures are calculated which disclose the differences and similarities between folder structures of the two user groups. The results shows that the group conducting more administrative activities has broader and shallower folders than the PhD group who performs more research activities, and the folders of the PhD group are more populated over deeper levels of the trees than those of the administrative group. The study improves our understanding of the various quantitative measures in investigating personal computer folder structures, and furthermore contributes to our knowledge of the information organization structure in personal information systems.
• Lee, J. H., Choi, K., Hu, X., & Stephen Downie, J. (2013). K-pop genres: A cross-cultural exploration. In 14th International Society for Music Information Retrieval Conference, ISMIR 2013.
  More info

  Current music genre research tends to focus heavily on classical and popular music from Western cultures. Few studies discuss the particular challenges and issues related to non-Western music. The objective of this study is to improve our understanding of how genres are used and perceived in different cultures. In particular, this study attempts to fill gaps in our understanding by examining K-pop music genres used in Korea and comparing them with genres used in North America. We provide background information on K-pop genres by analyzing 602 genre-related labels collected from eight major music distribution websites in Korea. In addition, we report upon a user study in which American and Korean users annotated genre information for 1894 K-pop songs in order to understand how their perceptions might differ or agree. The results show higher consistency among Korean users than American users demonstrated by the difference in Fleiss’ Kappa values and proportion of agreed genre labels. Asymmetric disagreements between Americans and Koreans on specific genres reveal some interesting differences in the perception of genres. Our findings provide some insights into challenges developers may face in creating global music services.
• Hu, X., & Kando, N. (2012). User-centered Measures Vs. System effectiveness in finding similar songs. In 13th International Society for Music Information Retrieval Conference, ISMIR 2012.
  More info

  User evaluation in the domain of Music Information Retrieval (MIR) has been very scarce, while algorithms and systems in MIR have been improving rapidly. With the maturity of system-centered evaluation in MIR, time is ripe for MIR evaluation to involve users. In this study, we compare user-centered measures to a system effectiveness measure on the task of retrieving similar songs. To collect user-centered measures, we conducted a user experiment with 50 participants using a set of music retrieval systems that have been evaluated by a system-centered approach in the Music Information Retrieval Evaluation eXchange (MIREX). The results reveal weak correlation between user-centered measures and system effectiveness. It is also found that user-centered measures can disclose difference between systems when there was no difference on system-effectiveness. © 2012 International Society for Music Information Retrieval.
• Hu, X., & Lee, J. H. (2012). A Cross-cultural study of music mood perception between American and Chinese listeners. In 13th International Society for Music Information Retrieval Conference, ISMIR 2012.
  More info

  Music mood has been recognized as an important access point for music and many online music services support browsing by mood. However, how people judge music mood has not been well studied in the Music Information Retrieval (MIR) domain. In particular, people's cultural background is often assumed to be an important factor in music mood perception, but this assumption has not been verified by empirical studies. This paper reports on a study comparing mood judgments on a set of 30 songs by American and Chinese people. Results show that mood judgments do indeed differ between American and Chinese respondents. Furthermore, respondents' mood judgments tended to agree more with other respondents from the same culture than those from the other group. Both the song characteristics (e.g., genre, lyrical or instrumental) and the non-cultural background of the respondents (e.g., age, gender, familiarity with the songs) were analyzed to further examine the difference in mood judgments. Findings of this study help further our understanding on how cultural background affects mood perception. Also discussed in this paper are implications for designing MIR systems for cross-cultural music mood classification and recommendation. © 2012 International Society for Music Information Retrieval.
• Lee, J. H., & Hu, X. (2012). Generating ground truth for music mood classification using mechanical turk. In 12th ACM/IEEE-CS Joint Conference on Digital Libraries, JCDL '12.
  More info

  Mood is an important access point in music digital libraries and online music repositories, but generating ground truth for evaluating various music mood classification algorithms is a challenging problem. This is because collecting enough human judgments is time-consuming and costly due to the subjectivity of music mood. In this study, we explore the viability of crowdsourcing music mood classification judgments using Amazon Mechanical Turk (MTurk). Specifically, we compare the mood classification judgments collected for the annual Music Information Retrieval Evaluation eXchange (MIREX) with judgments collected using MTurk. Our data show that the overall distribution of mood clusters and agreement rates from MIREX and MTurk were comparable. However, Turkers tended to agree less with the pre-labeled mood clusters than MIREX evaluators. The system evaluation results generated using both sets of data were mostly the same except for detecting one statistically significant pair using Friedman's test. We conclude that MTurk can potentially serve as a viable alternative for ground truth collection, with some reservation with regards to particular mood clusters. © 2012 ACM.
• Yang, Y. H., & Hu, X. (2012). Cross-cultural music mood classification: A comparison on English and Chinese songs. In 13th International Society for Music Information Retrieval Conference, ISMIR 2012.
  More info

  Most existing studies on music mood classification have been focusing on Western music while little research has investigated whether mood categories, audio features, and classification models developed from Western music are applicable to non-Western music. This paper attempts to answer this question through a comparative study on English and Chinese songs. Specifically, a set of Chinese pop songs were annotated using an existing mood taxonomy developed for English songs. Six sets of audio features commonly used on Western music (e.g., timbre, rhythm) were extracted from both Chinese and English songs, and mood classification performances based on these feature sets were compared. In addition, experiments were conducted to test the generalizability of classification models across English and Chinese songs. Results of this study shed light on cross-cultural applicability of research results on music mood classification. © 2012 International Society for Music Information Retrieval.
• Hu, X., & Yu, B. (2011). Exploring the relationship between mood and creativity in rock lyrics. In 12th International Society for Music Information Retrieval Conference, ISMIR 2011.
  More info

  The relationship between mood and creativity has been widely studied in psychology, however, no conclusion is reached in terms of which mood triggers high creativity, positive or negative. This paper provides new insights to this on-going argument by examining the relationship between lyrics creativity and music mood. We use three computational measures to gauge lyrics creativity: Type-to- Token Ratio, word norms fraction, and WordNet similarity. We then test three hypotheses regarding differences in lyrics creativity between music with different moods on 2715 U.S. rock songs. The three measures led to consistent findings that lyrics of negative and sad songs demonstrate higher linguistic creativity than those of positive and happy songs. Our findings support previous studies in psycholinguistics that people write more creatively when the text conveys sad or negative sentiment, and contradict previous research that positive mood triggers more unusual word associations. The result also indicates that different measures capture different aspects of lyrics creativity. © 2011 International Society for Music Information Retrieval.
• Cheng, J., Hu, X., & Heidorn, P. B. (2010). New measures for the evaluation of interactive information retrieval systems: Normalized task completion time and normalized user effectiveness. In Proceedings of the American Society for Information Science and Technology, 47.
  More info

  User satisfaction, though difficult to measure, is the main goal of Information Retrieval (IR) systems. In recent years, as Interactive Information Retrieval (IIR) systems have become increasingly popular, user effectiveness also has become critical in evaluating IIR systems. However, existing measures in IR evaluation are not particularly suitable for gauging user satisfaction and user effectiveness. In this paper, we propose two new measures to evaluate IIR systems, the Normalized Task Completion Time (NT) and the Normalized User Effectiveness (NUE). The two measures overcome limitations of existing measures and are efficient to calculate in that they do not need a large pool of search tasks. A user study was conducted to investigate the relationships between the two measures and the user satisfaction and effectiveness of a given IR system. The learning effects described by NT, NUE, and the task completion time were also studied and compared. The results show that NT is strongly correlated with user satisfaction, NUE is a better indicator of system effectiveness than task completion time, and both new measures are superior to task completion time in describing the learning effect of the given IR system.
• Hu, X., & Downie, J. S. (2010). Improving mood classification in music digital libraries by combining lyrics and audio. In 10th Annual Joint Conference on Digital Libraries, JCDL 2010.
  More info

  Mood is an emerging metadata type and access point in music digital libraries (MDL) and online music repositories. In this study, we present a comprehensive investigation of the usefulness of lyrics in music mood classification by evaluating and comparing a wide range of lyric text features including linguistic and text stylistic features. We then combine the best lyric features with features extracted from music audio using two fusion methods. The results show that combining lyrics and audio significantly outperformed systems using audio-only features. In addition, the examination of learning curves shows that the hybrid lyric + audio system needed fewer training samples to achieve the same or better classification accuracies than systems using lyrics or audio singularly. These experiments were conducted on a unique large-scale dataset of 5,296 songs (with both audio and lyrics for each) representing 18 mood categories derived from social tags. The findings push forward the state-of-the-art on lyric sentiment analysis and automatic music mood classification and will help make mood a practical access point in music digital libraries. © 2010 ACM.
• Hu, X., & Stephen Downie, J. (2010). When lyrics outperform audio for music mood classification: A feature analysis. In 11th International Society for Music Information Retrieval Conference, ISMIR 2010.
  More info

  This paper builds upon and extends previous work on multi-modal mood classification (i.e., combining audio and lyrics) by analyzing in-depth those feature types that have shown to provide statistically significant improvements in the classification of individual mood categories. The dataset used in this study comprises 5,296 songs (with lyrics and audio for each) divided into 18 mood categories derived from user-generated tags taken from last.fm. These 18 categories show remarkable consistency with the popular Russell's mood model. In seven categories, lyric features significantly outperformed audio spectral features. In one category only, audio outperformed all lyric features types. A fine grained analysis of the significant lyric feature types indicates a strong and obvious semantic association between extracted terms and the categories. No such obvious semantic linkages were evident in the case where audio spectral features proved superior. © 2010 International Society for Music Information Retrieval.
• Hu, X. (2009). Categorizing music mood in social context. In Proceedings of the American Society for Information Science and Technology, 46.
• Hu, X., Stephen Downie, J., & Ehmann, A. F. (2009). Lyric text mining in music mood classification. In 10th International Society for Music Information Retrieval Conference, ISMIR 2009.
  More info

  This research examines the role lyric text can play in improving audio music mood classification. A new method is proposed to build a large ground truth set of 5,585 songs and 18 mood categories based on social tags so as to reflect a realistic, user-centered perspective. A relatively complete set of lyric features and representation models were investigated. The best performing lyric feature set was also compared to a leading audio-based system. In combining lyric and audio sources, hybrid feature sets built with three different feature selection methods were also examined. The results show patterns at odds with findings in previous studies: audio features do not always outperform lyrics features, and combining lyrics and audio features can improve performance in many mood categories, but not all of them. © 2009 International Society for Music Information Retrieval.
• Hu, X., Downie, J. S., Laurier, C., Bay, M., & Ehmann, A. F. (2008). The 2007 mirex audio mood classification task: Lessons learned. In 9th International Conference on Music Information Retrieval, ISMIR 2008.
  More info

  Recent music information retrieval (MIR) research pays increasing attention to music classification based on moods expressed by music pieces. The first Audio Mood Classification (AMC) evaluation task was held in the 2007 running of the Music Information Retrieval Evaluation eXchange (MIREX). This paper describes important issues in setting up the task, including dataset construction and ground-truth labeling, and analyzes human assessments on the audio dataset, as well as system performances from various angles. Interesting findings include system performance differences with regard to mood clusters and the levels of agreement amongst human judgments regarding mood labeling. Based on these analyses, we summarize experiences learned from the first community scale evaluation of the AMC task and propose recommendations for future AMC and similar evaluation tasks.
• Hu, X., & Downie, J. S. (2007). Exploring mood metadata: Relationships with genre, artist and usage metadata. In 8th International Conference on Music Information Retrieval, ISMIR 2007.
  More info

  There is a growing interest in developing and then evaluating Music Information Retrieval (MIR) systems that can provide automated access to the mood dimension of music. Mood as a music access feature, however, is not well understood in that the terms used to describe it are not standardized and their application can be highly idiosyncratic. To better understand how we might develop methods for comprehensively developing and formally evaluating useful automated mood access techniques, we explore the relationships that mood has with genre, artist and usage metadata. Statistical analyses of term interactions across three metadata collections (AllMusicGuide.com, epinions.com and Last.fm) reveal important consistencies within the genre-mood and artist-mood relationships. These consistencies lead us to recommend a cluster-based approach that overcomes specific term-related problems by creating a relatively small set of data-derived "mood spaces" that could form the ground-truth for a proposed MIREX "Automated Mood Classification" task. ©2007 Austrian Computer Society (OCG).
• Hu, X., Bay, M., & Downie, J. S. (2007). Creating a simplified music mood classification ground-truth set. In 8th International Conference on Music Information Retrieval, ISMIR 2007.
  More info

  A standardized mood classification testbed is needed for formal cross-algorithm comparison and evaluation. In this poster, we present a simplification of the problems associated with developing a ground-truth set for the evaluation of mood-based Music Information Retrieval (MIR) systems. Using a dataset derived from Last.fm tags and the USPOP audio collection, we have applied a K-means clustering method to create a simple yet meaningful cluster-based set of high-level mood categories as well as a ground-truth dataset. ©2007 Austrian Computer Society (OCG).
• Wang, X., Zhai, C., Hu, X., & Sproat, R. (2007). Mining correlated bursty topic patterns from coordinated text streams. In KDD-2007: 13th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining.
  More info

  Previous work on text mining has almost exclusively focused on a single stream. However, we often have available multiple text streams indexed by the same set of time points (called coordinated text streams), which offer new opportunities for text mining. For example, when a major event happens, all the news articles published by different agencies in different languages tend to cover the same event for a certain period, exhibiting a correlated bursty topic pattern in all the news article streams. In general, mining correlated bursty topic patterns from coordinated text streams can reveal interesting latent associations or events behind these streams. In this paper, we define and study this novel text mining problem. We propose a general probabilistic algorithm which can effectively discover correlated bursty patterns and their bursty periods across text streams even if the streams have completely different vocabularies (e.g., English vs Chinese). Evaluation of the proposed method on a news data set and a literature data set shows that it can effectively discover quite meaningful topic patterns from both data sets: the patterns discovered from the news data set accurately reveal the major common events covered in the two streams of news articles (in English and Chinese, respectively), while the patterns discovered from two database publication streams match well with the major research paradigm shifts in database research. Since the proposed method is general and does not require the streams to share vocabulary, it can be applied to any coordinated text streams to discover correlated topic patterns that burst in multiple streams in the same period. © 2007 ACM.
• Downie, J. S., & Hu, X. (2006). Review mining for music digital libraries: Phase II. In 6th ACM/IEEE-CS Joint Conference on Digital Libraries 2006: Opening Information Horizons, JCDL '06, 2006.
  More info

  We continue our work on the automatic mining of user-created music reviews towards the goal of connecting user opinions to music objects in Music Digital Libraries (MDL). We demonstrate an experimental system which automatically discovered the key descriptive patterns that differentiated positive from negative reviews which helps us to better understand our successful Phase I results. Comparison to an earlier study indicates an important consistency across projects that warrants further investigation. Copyright 2006 ACM.
• Hu, X., Downie, J. S., & Ehmann, A. F. (2006). Exploiting recommended usage metadata: Exploratory analyses. In 7th International Conference on Music Information Retrieval, ISMIR 2006.
  More info

  In this paper, we conduct a series of exploratory analyses on the user-recommended usages of music as generated by 1,042 reviewers who have posted to www.epinions.com. Using hierarchical clustering methods on data derived from the co-occurrence analyses of usage and genre, usage and artist, and usage and album, we are able to conclude that further investigation of user-recommended usage metadata is warranted, especially with regard to its implications for future iterations of the Music Information Retrieval Evaluation eXchange (MIREX). © 2006 University of Victoria.
• Downie, J. S., Ehmann, A. F., & Hu, X. (2005). Music-to-Knowledge (M2K): A prototyping and evaluation environment for music digital library research. In 5th ACM/IEEE Joint Conference on Digital Libraries - Digital Libraries: Cyberinfrastructure for Research and Education.
• Hu, X., Downie, J. S., West, K., & Ehmann, A. (2005). Mining music reviews: Promising preliminary results. In 6th International Conference on Music Information Retrieval, ISMIR 2005.
  More info

  In this paper we present a system for the automatic mining of information from music reviews. We demonstrate a system which has the ability to automatically classify reviews according to the genre of the music reviewed and to predict the simple one-to-five star rating assigned to the music by the reviewer. This experiment is the first step in the development of a system to automatically mine arbitrary bodies of text, such as weblogs (blogs) for musically relevant information. © 2005 Queen Mary, University of London.
• Hu, X., Bandhakavi, S., & Zhai, C. (2003). Error Analysis of Difficult Trec Topics. In Proceedings of the 26th annual international ACM SIGIR conference on Research and development in information retrieval.
  More info

  Given the experimental nature of information retrieval, progress critically depends on analyzing the errors made by existing retrieval approaches and understanding their limitations. Our research explores various hypothesized reasons for hard topics in TREC-8 ad hoc task, and shows that the bad performance is partially due to the existence of highly distracting sub-collections that can dominate the overall performance.