4fcad3f4c547f8aa8d6421201c9db97

Welcome to the 2nd MSDS 2024!

The Centre for Data Sciences (CDS) of the University of Macau is excited to extend to you a warm welcome to showcase of progress of the rapidly growing and highly promising field of data sciences in the 2nd Macau Symposium on Data Science (MSDS 2024), which will be held in the University of Macau on 29 November 2024.

The aims of this symposium are to exchange on the advances and achievements in the field of data sciences, spanning from marketing, financial technology, artificial intelligence, law, education, language, precision medicine and smart governance etc. In particular, the event will bring together speakers from diverse backgrounds to discuss the latest research on data science or data science tools. 

The program will feature plenary sessions, keynote speakers and discussion on the following topics:

1) Data mining and data management;
2) Machine learning algorithms and techniques;
3) Natural language processing;
4) Big data analytics and processing;
5) Data visualization and storytelling;
6) Ethical and legal issues in data science;
7) Applications of data science in various industries.

Schedule

 From ToSpeaker 講者
09:1509:25Opening Remarks
09:2509:30Group Photo
09:3010:15KEYNOTE Presentation, Prof. Huaqing HONG
Transforming Teaching & Learning with Big-data Learning Analytics
10:1510:45

Industrial Talk – Huawei

Huawei - Wikipedia

10:4511:05Coffee break
11:0511:50KEYNOTE Presentation, Prof. Sourav S. BHOWMICK
Data-driven COI Management: From Socio-technical Data Systems Research to Impact
11:5012:20Technical Session 1, Prof. Jiangtao WANG
Incorporating Domain Knowledge to Mitigate Low-Data Challenge in AI for Health: Vision and Case Studies
12:2012:50Industrial Talk – Nvidia
https://www.nvidia.com/content/nvidiaGDC/us/en_US/about-nvidia/legal-info/logo-brand-usage/_jcr_content/root/responsivegrid/nv_container_392921705/nv_container_412055486/nv_image.coreimg.100.1070.png/1703060329095/nvidia-logo-horz.png
12:5014:20Lunch break
14:2015:05KEYNOTE Presentation, Prof. Yanzhi Ll
Data-driven Decision Making Powered by Insights
15:0515:35Technical Session 2, Prof. Ben KAO
AI-assisted Community Legal Information Access
15:3516:05Technical Session 2, Prof. Jianhui XIE
Structure Identification for Partially Linear Partially Concave Models
16:0516:35Technical Session 2, Prof. Li CHEN
A User-Centric Approach to Developing Trustworthy Recommender Systems
16:3516:55Coffee break
16:5517:25Technical Session 3, Prof. Rustam SHADIEV
AI in Education: Enhancing Learning through Speech-Enabled Technologies
17:2517:55Technical Session 3, Prof. Yi DING
Learning the Stochastic Discount Factor
18:0019:30Dinner for invited guest, 萬豪軒

Keynotes

Sourav S. BHOWMICK
 

Sourav S. Bhowmick is an Associate Professor in the College of Computing & Data Science (CCDS),
Nanyang Technological University, Singapore. His research interests are in data management, human-data
interaction, and data analytics. His research has appeared in premium venues such as ACM SIGMOD,
VLDB, and VLDB Journal. He is co-recipient of Best Paper Awards in ACM CIKM 2004, ACM BCB
2011, VLDB 2021, and ER 2023. Sourav is serving as a member of the SIGMOD Executive Committee,
SIGMOD Awards Committee, a regular member of the PVLDB advisory board, and an elected trustee
of the VLDB Endowment. He is a co-recipient of several service awards including VLDB Service Award
in 2018, Distinguished AE Award in SIGMOD 2021, SIGMOD 2023 and VLDB 2022, and Distinguished
Reviewer Award in VLDB 2020 and VLDB 2023. He was inducted into Distinguished Members of the
ACM in 2020. Sourav is a strong advocate of research that directly or indirectly impacts end users.

Upholding high standards of fairness is central to review processes in peer-reviewed conferences. To this end, many aspects of review and selection processes have been strengthened over the past 50 years, but the ways we declare and manage conflicts of interest (COI) have changed little if at all. In this talk, we present our experience in building and deploying a novel socio-technical data system called CLOSET for detecting and managing COIs in peer-reviewed conferences. We highlight the intriguing and unconventional socio-technical data challenges that lurk underneath and how we mitigate them. We end the talk by highlighting the practical impact of CLOSET in the data management community and an array of unconventional open research challenges at the intersection of psychology, data management, and data analytics.

Huaqing HONG
洪化清

Dr Hong holds a MA and a PhD on Corpus Linguistics from the National University of Singapore. He was formerly a Senior Assistant Director and Head of e-Learning Research & Development at the Lee Kong Chian School of Medicine, Nanyang Technological University. He is currently a Professor and the Director of the Centre for Language Intelligence and Smart Education at the Shanghai International Studies University.

Despite the proliferation of educational technologies, fundamental pedagogical practices have remained largely unchanged. This presentation examines the critical role of big-data learning analytics in catalyzing genuine educational transformation, focusing on evidence-based pedagogy and its potential to enhance decision-making in teaching and learning. We will explore how big-data analytics can inform adaptive learning systems, enable proactive interventions, and facilitate evidence-based curriculum design. The discourse will address the current landscape of educational technology, the importance of evidence-based pedagogy, the integration of big-data learning analytics, and the significance of practical implementation strategies in classroom settings. Emphasizing the pivotal role of data science in fulfilling the promise of educational technology, we will present case studies from Nanyang Technological University to illustrate real-world applications and outcomes. This talk aims to bridge the gap between theoretical potential and practical application, offering insights into how educators and institutions can effectively leverage big-data learning analytics to realize the long-promised transformation of teaching and learning, while acknowledging the challenges and necessary considerations for successful implementation.

Yanzhi LI
李彥志

David Li is Professor of Marketing and Management Sciences and an Affiliated Professor of the School of Data Science at the City University of Hong Kong. He is also Head of the Department of Marketing and Director of Fintech and Business Analytics Centre at CityUHK. He received a bachelor’s degree in Computer Science from Tsinghua and his Ph.D. in Industrial Engineering and Engineering Management from HKUST. David develops data-driven decision-making methodologies for various industries and has provided consulting services to firms in the Greater Chain region. He co-founded Pricing Service, Inc., a start-up specializing in hotel revenue management.
Decision making is central to business operations, and ultimately decision making starts from data, big or small. We discuss several possible paths from data to decisions, with example applications. Finally, we focus on the problem of personalized dynamic promotion for storable goods. We show how we can dramatically improve the performance of offline reinforcement learning by incorporating  domain knowledge and analytical properties into the algorithmic framework. We also compare the proposed approach with other existing approaches and illustrate that theoretically appealing approaches may lead to no business gains in a data-limited environment.

Invited Talks

Li CHEN
陳黎

Prof. Li Chen is currently a Professor and Associate Head (Research) in the Department of Computer Science at Hong Kong Baptist University (HKBU). She obtained her PhD degree in Computer Science from the Swiss Federal Institute of Technology in Lausanne (EPFL), Switzerland, and Bachelor and Master degrees from Peking University, China. Her recent research focus has mainly been on online decision support, data-driven personalization, and recommender systems, with applications covering various domains including entertainment, digital media, education, e-commerce, and psychological well-being. She has authored and co-authored over 150 publications, with 10,400 citations so far (H-index 49). 

 

Recommender systems have become increasingly prevalent across various online platforms (e.g., e-commerce, entertainment, and social media), delivering personalized information and services tailored to individual users. While numerous studies have concentrated on enhancing recommendation
accuracy to engage users, there is a pressing need for a more user-centric approach that prioritizes practical benefits for users. In this talk, I will present a user-centric methodology for effectively utilizing recommender systems. Specifically, I will discuss two case studies. The first case study
highlights the integration of tradeoff decision support within recommender
systems to improve users’ decision-making quality in high-stakes product domains.

The second case study addresses the filter bubble phenomenon by considering users’ personal characteristics, such as personality traits, to develop personalized diversity strategies. Both studies demonstrate the significant impact of these design factors on enhancing users’ trust in the system. The findings have the potential to advance research on trustworthy and responsible recommender systems from the users’ perspective.

Yi DING
丁一

Prof Ding got Ph.D in Business Statistics at Hong Kong University of Science and Technology in July 2020. She is Assistant Professor in Business Economics at the Faculty of Business Administration of University of Macau. Her research interests include Financial econometrics; High-dimensional statistics; Financial technology; Statistical learning; Portfolio optimization; Asset allocation; High-frequency financial data.

We develop a statistical framework to learn the high-dimensional stochastic dis- count factor (SDF) from a large set of characteristic-based portfolios. Specifically, we use the MAXSER method proposed in Ao et al. (2019) to screen for potentially useful factors, and develop a statistical inference theory for further factor selection and final SDF portfolio construction. Applying our approach to 194 characteristic-based portfolios, we find that the SDF constructed by about 20 of them performs well in achieving a high Sharpe ratio and explaining the cross-section of expected returns of various portfolios.

 

Ben KAO
高志明

Ben Kao received the B.Sc. degree in computer science from the University of Hong Kong in 1989 and the Ph.D. degree in computer science from Princeton University in 1995. From 1992 to 1995, he was a research fellow at Stanford University. He is currently full professor in the School of Computing and Data Science (CDS) at The University of Hong Kong, associate head of The Innovation Academy, the chairperson of The Board of Faculty of Engineering, and HKU Senate member. His research interests include database management systems, data mining, information retrieval systems, AI and natural language processing.

In the contemporary era, legal information, encompassing court judgments and legislation, is generally accessible online in numerous countries. However, the online availability of this information does not necessarily equate to effective public access to legal knowledge. It presents significant challenges for individuals without legal expertise to acquire legal knowledge due to two primary reasons. Firstly, the online content predominantly consists of primary legal sources, such as cases and statutes, which are typically written in formal legal terminology that can be challenging for the general public to comprehend. Secondly, the public may lack knowledge of the applicable legal principles in their specific legal situations. Given the vast number of documents, it becomes arduous for users to identify the relevant legal sources when seeking solutions to their legal challenges. In this presentation, we will demonstrate several AI tools that we have integrated into our online legal information platforms, specifically HKLII and CLIC. We will elucidate how these tools facilitate enhanced public access to legal information.

Shadiev RUSTAM
 

Professor Rustam Shadiev earned his Ph.D. in Network Learning Technology from National Central University, Taiwan, China in 2012. He is currently a tenured professor at the College of Education, Zhejiang University, China, specializing in advanced learning technologies for language learning and cross-cultural education. Professor Shadiev is a Fellow of the British Computer Society and a Senior Member of the Institute of Electrical and Electronics Engineers (IEEE). In 2019, he was honored with the title of Distinguished Professor of Jiangsu Province, China. Additionally, he has been recognized as one of the most cited Chinese researchers in the field of education by Elsevier, Scopus, and Shanghai Ranking for four consecutive years, from 2020 to 2023. Professor Shadiev was also listed among Stanford/Elsevier’s Top 2% Scientists in the field of education and educational research in 2023 and 2024.

This presentation explores the transformative impact of artificial intelligence in education, with a focus on speech-enabled technologies such as Speech-to-Text Recognition (STR), Speech-Enabled Corrective Feedback (SECF), Speech-Enabled Language Translation (SELT), and speech-enabled ChatGPT. Drawing on our empirical studies, the talk examines how these AI-driven tools enhance learning outcomes, reduce cognitive load, alleviate learning anxiety, and boost student confidence.

The first study investigates the use of STR technology in lectures, demonstrating its effectiveness in improving learning performance and reducing cognitive load, particularly for students with lower English proficiency. The second study introduces SECF, which combines STR with real-time corrective feedback to enhance speaking skills in English as a foreign language (EFL). The results show significant improvements in pronunciation, grammar, and discourse, along with reduced anxiety and increased confidence. The third study explores SELT technology, where real-time translation aids communication between students from different cultural backgrounds, speaking in their native languages. Findings reveal enhanced cross-cultural understanding and communication competence. Lastly, the fourth study examines the use of speech-enabled ChatGPT, where students interact with AI peers to deepen their understanding of learning content. Results indicate that students received timely, credible responses to their questions and felt supported throughout the learning process. Through these four studies, this presentation highlights the potential of AI to reshape learning environments and provides practical insights for educators and researchers interested in integrating AI technologies into their teaching practices.

Jiangtao WANG
王江濤

Dr. Jiangtao Wang is a Professor at the School of AI and Data Science, University of Science and Technology of China (USTC). Before joining USTC, he was an Assistant Professor and Associate Professor in the UK. Dr. Wang specializes in developing AI algorithms to analyze complex healthcare data, aiming to improve healthcare delivery. He has created advanced machine learning models for population health monitoring, diagnosis prediction, drug-drug interaction analysis, and COVID-19 severity estimation, often outperforming current methods on real-world data. In 2021, Dr. Wang received the EPSRC New Investigator Award, and in 2023, he was selected for the UK Future Leader Fellow development network.

Artificial Intelligence (AI) is transforming healthcare by enabling predictive models, automated diagnostics, and personalized treatment plans. However, realizing the full potential of AI in health faces a significant challenge: the scarcity of high-quality labeled data. Many healthcare applications, such as disease prediction, drug interaction analysis, and public health monitoring, require vast amounts of data to train accurate models. Yet, health data is often limited, fragmented, or difficult to obtain due to privacy concerns, patient diversity, and the complexity of medical conditions. This “low-data challenge” poses a critical barrier to advancing AI-driven solutions in healthcare. In this talk, I explore how domain knowledge—such as medical expertise, clinical guidelines, and epidemiological insights—can play a crucial role in addressing this issue. By incorporating expert knowledge into AI models, we can enhance learning from small datasets and generate more reliable predictions. Through case studies in population health, we will demonstrate practical applications, such as the use of expert-informed models for predicting health outcomes and assessing environmental health risks. These examples showcase how combining AI techniques with domain knowledge not only helps overcome the data limitation but also produces actionable insights for real-world health challenges.

Jianhui XIE
謝建輝

Jianhui Xie is an associate professor in International School of Business & Finance at SunYat-sen University. He received his PhD in management science and engineering from the University of Science and Technology of China. His research interests include productivity and efficiency analysis, shape constrained regression, and nonparametric statistics. His research papers have appeared in journals such as Journal of Business & Economic Statistics, European Journal of Operational Research, and Omega.

Partially linear partially concave models are semiparametric regression models that can capture linear and concavity-constrained nonlinear effects within one framework. A fundamental problem of this kind of model is deciding which covariates have linear effects and which covariates have strictly concave effects. Assuming that the true regression function is partially linear partially concave and sparse, we develop two structure selection procedures for classifying the covariates into linear, strictly concave, and irrelevant subsets. We show that the procedures based on penalized concavity-constrained additive regressions can correctly identify structures even if the underlying true functions are nonadditive; namely, the proposed procedures are additively faithful in a general setting. We prove that consistent structure selection is achievable when the total number of covariates and the number of concave covariates grow at polynomial rates with sample size. We introduce algorithms to implement the proposed procedures and demonstrate their performance by simulation analysis.

Committee

Chair

u-leong-hou-ryan-1-275x300

Prof. Leong Hou U, FST

余亮豪 教授

Members

dingqi

Prof. Dingqi YANG, FST

楊丁奇 教授

cherisc

Prof. Cheris Wing Chi CHOW, FBA

周詠芝 教授

henrylei

Prof. Henry Chun Kwok LEI, FBA

李振國 教授

muruga-perumal-ramaswamy

Prof. Muruga Perumal RAMASWAMY, FLL

terence-chuen-wai-poon-819x1024-1-800x1000

Prof. Chun Wai POON, FHS

潘全威 教授

Barry Reynolds

Prof. Barry Lee REYNOLDS, FED

雷貝利 教授

tjcai_c-800x976

Prof. Tianji CAI, FSS

蔡天驥 教授

lisongqing

Prof. Songqing LI, FAH

勵松青 教授

Venue

Date: 29  November, 2024

Address: Ground floor Function Hall,  N1 Building – University of Macau

Map: UM-Map

About UM

About CDS (UM)

返回頂端