Data science has\nbecome the backbone of today\u2019s intelligent world. From data acquisition and\nanalysis to real-world deployment and transformation, the discipline is\nincreasingly embedded across every sector of society. In response to this\naccelerating momentum, the University of Macau (UM) established the Centre for\nData Science (CDS) as an interdisciplinary platform dedicated to advancing\ndata-driven research and innovation. Through cross-disciplinary collaboration,\nUM brings together expertise spanning the entire data science pipeline and is\ncommitted to fostering a more sustainable and intelligent future.<\/p>\n\n
Uncovering Patterns\nThrough Large-Scale Data<\/p>\n\n
Data is a critical\nresource driving social development, and data mining is the key technology that\nunlocks its value. Through advanced algorithms and modelling techniques,\nresearchers can move beyond surface\u2011level information to uncover hidden\npatterns, reshaping how complex social systems are understood and analysed. A\nresearch team led by Cai Tianji, associate dean of UM\u2019s Faculty of Social\nSciences and professor in CDS, has applied data mining techniques to examine\njudicial decision\u2011making using a vast corpus of court\njudgments.<\/p>\n\n
Since 2015, Prof\nCai\u2019s team has extracted publicly available effective judgments\u2014excluding cases\nunsuitable for online publication\u2014from China\u2019s People\u2019s Courts at all levels\nvia the China Judgments Online platform. To date, team members have collected\nmore than 20 million judgments and conducted large-scale, systematic analyses\nto examine factors influencing sentencing and identify patterns of crime\nbehaviour across China. Drawing on this extensive dataset, team members have\npublished more than ten papers in leading international journals, including Journal\nof Quantitative Criminology <\/i>and the International Journal of Drug Policy<\/i>,\ncovering topics such as human trafficking, drug offences, and sexual assault.\nThese studies provide quantitative insights into sentencing practices across\ndifferent categories of criminal cases.<\/p>\n\n
For example, the\npaper \u2018Paying Money for Freedom: Effects of Monetary Compensation on Sentencing\nfor Criminal Traffic Offenses\u2019, co-authored by Prof Cai and his former doctoral\nstudent Xin Yanyu (now a lecturer at the Research Institute of Social\nDevelopment, Southwestern University of Finance and Economics) uses data mining\nto obtain over 140,000 valid sentencing documents as the basis for analysis.\nEmploying a joint modelling approach that considers both sentence length and\nprobation as outcomes, the study adopts a zero-truncated generalised inflated\nPoisson model to address distributional characteristics of sentence length,\nsuch as the absence of zero values and the concentration of observations at\nspecific points, thereby revealing the potentially significant impact of\nmonetary compensation on sentencing in traffic accident cases.<\/p>\n\n
Transitioning from\nan era of \u2018data scarcity\u2019 to one of data-driven research has not been without\nchallenges. Following several redesigns of the China Judgments Online platform,\nthe team has continuously refined its web-scraping code and data extraction\nstrategies to improve efficiency while ensuring legal compliance. To date, they\nhave collected over 10 terabytes of data and upgraded their infrastructure from\ntraditional hard drives to a more secure and scalable Network Attached Storage\n(NAS) system. Extracting meaningful structured information from massive volumes\nof unstructured legal text presents an additional technical challenge. For\nexample, identifying prison terms requires the application of Named Entity\nRecognition (NER) techniques, followed by text processing and value\nnormalisation to accurately extract sentencing information.<\/p>\n\n
Benefiting from\nUM\u2019s open and collaborative research environment, the team has continued to\nrefine its methodologies through academic exchange. Members are now exploring\nthe use of advanced artificial intelligence (AI) technologies to further\nenhance text extraction, moving closer to their goal of building a\ncomprehensive and reliable sentencing database to support future research.<\/p>\n\n
Transforming\nLarge-Scale Data Into Intelligent Decisions<\/p>\n\n
At the intersection\nof social development and technological innovation, large-scale data processing\nprovides an evidence-based foundation for collective decision-making and\nadvanced research. By enabling data to be measured, modelled, and continuously\nrefined, data processing expands the boundaries of human understanding. <\/p>\n\n
A research team led\nby U Leong Hou, head of CDS, focuses on frontier challenges in large-scale data\nprocessing and intelligent decision-making. The team develops efficient methods\nfor analysing and optimising structured, spatiotemporal, and graph\u2011structured\ndata, while addressing scalability and data-efficiency bottlenecks in\nreinforcement learning for high-dimensional and dynamic environments. One\nrepresentative contribution is LIBKDV, a kernel density visualisation library\nthat enables high-resolution geospatial analytics on large-scale datasets.<\/p>\n\n
The team has\nintegrated this work with intelligent decision-making technologies for\ndata-intensive scenarios, including traffic signal control, transportation\nsystems, and resource allocation. By reducing data requirements and\ncomputational costs, the project accelerates the deployment of reinforcement\nlearning systems in real-world settings, contributing to advances in smart\nmobility and urban governance.<\/p>\n\n
At the academic\nlevel, the team has made substantial contributions to algorithm design,\ntheoretical analysis, and scalable system architecture, thereby establishing a\nforward-looking technical foundation for data-intensive intelligent\ndecision-making systems. The team\u2019s work has been recognised at leading\ninternational AI conferences, with multiple papers selected for spotlight\ndistinction and oral presentation. Recent studies further extend to key topics\nsuch as cross-scenario generalisation and large-scale real\u2011world\ndeployment.<\/p>\n\n
The project has\nprogressed through a multi-pronged strategy addressing two major challenges.\nThe first is the high demand for computational resources. The team prioritises\ndata-efficient and computation-aware algorithm design to reduce resource\nconsumption while maintaining performance. It has also developed reusable\nsoftware frameworks and modular components to enhance research productivity and\nmanage engineering complexity. The second challenge is the shortage of\ninterdisciplinary engineering talent. To address this, the team strengthens\ninterdisciplinary training to cultivate researchers capable of bridging theory,\nsystems, and implementation.<\/p>\n\n
Prof U emphasises\nthat UM\u2019s institutional support\u2014including sustained research funding, strong\ndoctoral training programmes, and the advanced GPU infrastructure of the Super\nIntelligent Computing Centre\u2014has been instrumental in the project\u2019s progress.\nLooking ahead, the team will continue advancing the \u2018AI for Data\u2019 paradigm,\nintegrating reinforcement learning with real-world data systems to optimise\nadaptive control strategies. A key focus will be traffic trend analysis and\nintelligent signal control in Macao, alongside broader validation of\ndata-centric reinforcement learning frameworks in complex, dynamic\nenvironments.<\/p>\n\n
Advancing\nMultilingual Intelligence Through Data Science<\/p>\n\n
From academic\nscholarship to the preservation of ancient tribal songs, from the\nstandardisation of multilingual medical research data to language mediation at\nemergency rescue sites, machine translation is quietly transforming the flow of\nknowledge and cross-cultural communication. Wong Fai, head of UM\u2019s Natural\nLanguage Processing & Portuguese\u2011Chinese Machine Translation Laboratory (NLP2<\/sup>CT)\nand professor in CDS, draws on his expertise in computational linguistics and\nnatural language processing to lead research focused on overcoming the\nchallenges of translating low-resource languages in response to Macao\u2019s\nmultilingual needs.<\/p>\n\n Leveraging the\nlaboratory\u2019s research strengths, the team has developed Macao\u2019s first\nChinese-Portuguese machine translation system, PCT, as well as the online\nChinese-Portuguese-English computer-aided translation platform (UM-CAT), which\nincorporates Macao\u2019s linguistic and cultural context. These systems have been\nadopted by various departments of the Macao SAR government for official\ndocument translation. Moreover, these technological innovations received the\nMacao Science and Technology Award in 2012 and 2022. In addition, UTran-i\nTechnology Ltd., founded by graduates of NLP2<\/sup>CT and incubated by\nUM\u2019s Centre for Innovation and Entrepreneurship, has developed a range of\nmultilingual products and services in Portuguese, English, Mandarin, and\nCantonese based on independently developed core technologies.<\/p>\n\n Behind these\nachievements lie more than 20 years of sustained research and development by\nProf Wong and his team. As he explains, Macao, as a multilingual yet relatively\nlow-resource environment, lacks large\u2011scale, high-quality parallel corpora for\nlanguage pairs such as Chinese-Portuguese and Cantonese\u2011Mandarin.\nIn addition, different professional domains exhibit fragmented terminology,\nstylistic variation, and distinct pragmatic conventions, all of which increase\nthe complexity of translation tasks. Real-world applications further demand\nintegration with multilingual, multimodal, and domain-specific knowledge bases,\nraising the technical threshold for system development. In this context, UM\u2019s\nInstitute of Collaborative Innovation promotes interdisciplinary collaboration\nacross computer science, linguistics, and data science, enabling the team to\nmitigate data scarcity and expand the scope of its research. Coordinated\ninternal and external research funding mechanisms have also provided sustained\nfinancial support for the team.<\/p>\n\n As language demands\nevolve, the team remains committed to positioning machine translation as a\nbridge for sharing Chinese culture with the world. This vision is reflected in\nthe translation pipeline developed specifically for Chinese literature, which\nwon first place in the literary translation task at the Conference on Machine\nTranslation (WMT 2024). Looking ahead, NLP2<\/sup>CT will further leverage\nits expertise to develop high-precision detection tools for AI-generated\ncontent. By addressing emerging concerns about AI credibility, the team aims to\npromote the responsible application of AI in scientific research, education,\nand cultural tourism.<\/p>\n\n Innovation Beyond\nConstraints<\/p>\n\n Extracting patterns\nfrom vast datasets, enabling informed decisions from complex information, and\ncreating opportunities within resource constraints require not only technological innovation\nbut also vision and perseverance. Through forward\u2011looking research and sustained\ninterdisciplinary collaboration, UM contributes to sustainable social\ndevelopment in the era of data intelligence.<\/p> <\/p> Text: Bell Leong, UM reporter Zhang Tongyu<\/p> Photo: Editorial board, with some provided by the interviewees<\/p> English Translation: Bess Che<\/p> Source: UMagazine<\/a><\/i> Issue 33<\/p> <\/p> All articles in Celebrating UM\u2019s 45th<\/sup>\nAnniversary: Research Powering the Future<\/i>:<\/a><\/p> Rector\u2019s\nWords<\/a><\/p> Advancing the Scientific Foundations\nof Chinese Medicine<\/a><\/p> Three Decades of Dedication to Chip\nInnovation and Real-World Impact<\/a><\/p> Connecting Everything to Build\nSmarter Cities<\/a><\/p> Opening New Pathways to Personalised\nHealth<\/a><\/p> Driving Breakthroughs in Materials\nEngineering<\/a><\/p> Safeguarding Sustainable Marine\nDevelopment Through Research<\/a><\/p> Shaping a Safe and Intelligent\nFuture<\/a><\/p> Advancing the Frontiers of Brain\nScience and Human Cognition<\/a><\/p> Promoting Social Development Through\nInterdisciplinary Research<\/a><\/p> Blueprint and Vision for the Future<\/a><\/p> Appendix in Celebrating UM\u2019s 45th<\/sup> Anniversary:\nResearch Powering the Future<\/i>:<\/p>