DS&E in HKUDS&E in HKU

HKU Musketeers Foundation Institute of Data Science

The HKU Musketeers Foundation Institute of Data Science (IDS) is the first of the ten research institutes in the University’s technological development blueprint for the next decade. Drawing on strengths in varied disciplines, the Institute aspires to connect our expertise and a University-wide interest in establishing critical mass in key areas, and to explore frontier research and applications in data science, computing, mathematics, and statistics. Researchers and their research postgraduate students recruited under the Data Science Cluster in the HKU-100 Recruitment Campaign launched in 2021 will be jointly appointed by both the Institute and various Departments from STEM and non-STEM disciplines.

R&D projects

The Department of Computer Science at HKU brings together a team of professionals dedicated to the development of data science and engineering through research, teaching, and knowledge exchanges. The Data Science & Engineering research group strives to provide a platform for exchanging research insights and nurturing future big data professionals through various seminars and workshops. The following are some highlights of the research projects and representative work that our departmental members are involved in.

Radiograph Representation by AI Algorithm

A research team led by Prof. Yizhoi Yu under the Department of Computer Science has developed a new AI-enabled medical image diagnosis that helps support medical professionals in reducing their workload and improving the diagnostic efficiency and accuracy. The new approach called ”REFERS” (Reviewing Free-texts Reports for Supervision) allows automatic acquisition of supervision signals from thousands of radiology reports at the same time, cutting down 90% of the original labour cost. With the help of AI technologies, REFERS helps with the generation of radiographic reports as well as radiograph-report matching. The team’s effort marks a breakthrough in realising a more generalised use of artificial intelligence in the medical field.

HINCare app

HINCare is an AI platform developed by Professor Reynold C.K. Cheng. It provides user friendly interface for elder care users to manage their timebanking account, a system where volunteers can earn time credits by performing a service (e.g., taking an elder to a clinic), which can then be used in exchange for services provided by another volunteer. With the help of the app’s Heterogeneous Information Network and AI technologies, participating NGOs will be able to use intelligent matching to find the best suited helper for each task. The platform also supports mobile app, which includes functions such as requesting for services and applying for volunteering opportunities. It has been selected for the Faculty Knowledge Exchange Award 2021 and Asia Smart Apps Award in recognition of Professor Cheng’s outstanding accomplishment.

Prediction Model for Illegal Drug Trafficking Sentencing

The HKU’s research team led by Professor Ben Kao of the Department of Computer Science and Professor Anne Cheung of the Faculty of Law has developed a computer program that can predict sentences in drug-trafficking cases with up to 99 percent accuracy. The software is dedicated to both professionals and those who have not received any legal training to effectively locate relevant legal information so that they can explain the legal principles to drug-offenders. Using AI technologies, the platform is taught to read more than 3,000 judgements on drug-trafficking cases by scanning through past court rulings and look for cases with similar circumstances to determine how long a sentence is likely to be. The sentencing prediction program can be assessed by the public at http://wwwnew2.hklii.hk/predictor.

Seminars

The Department of Computer Science has held various seminars to discuss how data manipulation can empower users in different aspects.

“The Power of Data Analytics and AI Techniques in the Digital Sector”

A talk on how data analytics and AI Techniques are used in the digital sector, and the current and future changes the industry is facing. Mr. Alan Chan (Executive Vice President, Lazada) also shared some tips on how to start a career in data and analytics.

“Efficient, Consistent and Secure Global-Scale Data Management”

This talk explored various technologies that aimed to facilitate building global-scale and edge-aware data management systems.

“Design Visualization for the Data Science Workflow”

Dr. Jian Zhao explains his research in addressing questions related to data visualization and discusses how visualization aids users in each of the three main stages of solving a practical data problem, including exploratory analysis, model curation, and insight communication. He also demonstrated a few interesting future directions of optimizing the above-mentioned workflow using interactive visualization.

Visit here for past seminars.