Research Interests

Big Data Analytics, Healthcare Analysis, Data Visualization, Human Computer Interaction, Explainable AI , Social Media , Social Network/Graph , Fintech Ecosystem. ​


Behaviour Modelling and Social Computing

Behaviour modelling and social computing are two fundamental research directions. Behaviour modelling is to derive user behavioural patterns, preferences, profiles from user behaviours and user generated textual data, while social computing is to analyse the social characteristics and trends demonstrated from intra-human or human-computer interactions.

Visual Data Analytics

Visualization and visual analytics have been introduced both in academia and industry: 1) to provide a clear view of users diverse behavior, transactions monitoring, premium fluctuations, and in complex everyday decision-making, 2) to characterize data, user and task, and 3) discovering imbalances and monitoring risk.

From our research group, we continuously publish research papers on these two topics and apply the developed techniques to help our industry partners with their business problems. Examples including anti-selection detection (Zurich One-Path), spam behaviour detection (Toutiao), and product/action recommendations (Yozo, Specifically, our strengths include ​

    • Identifying misbehaviours or abnormal behaviours from a population
    • Discovering the underlying behaviour patterns in a population
    • Design a new visual analytics solutions (VAS)
    • Recommending actions/items based on the underlying patterns
    • Techniques: topic modelling, representation learning and embedding, rule mining, sequential pattern mining, anomaly detection, time series analysis.


  • Discovering dynamic adverse behavior of users

Adverse selection (AS) is one of the significant causes of market failure worldwide. Analysis and deep insights into the Australian life insurance market show the existence of adverse activities to gain financial benefits, resulting in loss to insurance companies. Understanding the behavior of policyholders is essential to improve business strategies and overcome fraudulent claims. However, policyholders' behavior analysis is a complex process, usually involving several factors depending on their preferences and the nature of data such as data which is missing useful private information, the presence of asymmetric information of policyholders, the existence of anomalous information at the cell level rather than the data instance level and a lack of quantitative research. This project aims to analyze the life insurance policyholder's behavior to identify adverse behavior (AB).

Source: Islam, Md Rafiqul, Shaowu Liu, Rhys Biddle, Imran Razzak, Xianzhi Wang, Peter Tilocca, and Guandong Xu. "Discovering dynamic adverse behavior of policyholders in the life insurance industry." Technological Forecasting and Social Change 163 (2020): 120486.

  • Natural Language Interaction Aided with Data Visualization for Exploring Claim and Risk Management

Analysis of claims and risk management is the key task to avoid frauds and to provide risk management in the life insurance industry. Though visualization plays a fundamental role in supporting analysis tasks of the business domain, analyzing user behaviour remains a challenging task. The prevalence of natural language interactions aided with data visualization has become quite the norm. With the increasing demand of visualization tools and varying level of user expertise, it comes as no surprise the use of natural languages interface. However, the design of visual analytics tools aided with natural language interfaces (NLIs) for risk management and claim analysis requires thorough task analysis and domain expertise. In this project, our aims to design an alternative approach through a natural language interaction based interactive representation such as chart, pie, and histogram, which can be applied for investigating insurance claims and risk management.

  • Adaptive Deep Learning Underwriting Quality Assurance Modelling ​

The goal of this project is to provide a deep learning-based underwriting quality assurance model with long and short-term memory states. The model is expected to adapt to various external factors that influence the way underwriting evolves over time as well as should be able to depict and respond to the macro factors influence any changes in the underwriting policies and procedure. The deep learning-based model is expected to give higher weightage to recent events compared to lower weightage to older events, thus reducing underwriting workload and claim risk. This model may also be trained through reinforcement learning, for better results of solving exclusions, loading & policy outcome decisions based on large and complex disclosures.