UTS site search

Dr Wei Liu


Dr Wei Liu is a Data Science Research Program Leader and a Lecturer at the Advanced Analytics Institute, Faculty of Engineering and IT, the University of Technology Sydney (UTS). Before joining UTS, he was a Data Mining Research Fellow at the University of Melbourne, and then an Industry-focused Machine Learning Researcher and Project Manager working in the transportation industry at National ICT Australia (NICTA). He obtained his PhD degree in Data Mining Research from the University of Sydney (USYD).

His research outputs are mostly published in journals and conferences that are ranked at A* and A (i.e., top-prestige) by ARC ERA 2010 ranking and by Core academic ranking. He has received 3 Best Paper Awards.

Dr Liu is interested in industry-driven data analytics research that makes real-world impact. He has led a number of significant research projects funded by government agencies and industrial organisations, spanning internet security, insurance, trading, transportation, and infrastructure sectors. He has developed advanced data mining models and software tools for the transport industry, which accurately identify causes of road incidents. He has also designed cutting-edge predictive models for problems including rare event prediction, fraud/intrusion detection, emerging trends detection, etc. Details of some of his research projects are in the below:

  • "Analytics Model to Support Strategic Planning in a Regulatory Environment", industry partner: NSW Fair Trading; April - July 2015.
  • "Transport Data Science and Advanced Analytics", industry partner: National ICT Australia; July 2015 – June 2017.
  • “Traffic Watch for Transport Control Service”, industry partner: Transport Management Centre; May 2013 – June 2014.
  • “Congestion Propagation and Hotspot Detection in Sydney CBD”, industry partner: NSW RMS; Aug – Dec 2013.
  • “Data Fusion Technologies for Comprehensive Transport Data Analysis in Melbourne”, industry partner: VicRoads; Jun – Sep 2013.
  • “Time of Arrival Estimations using HD Vehicle Trajectories”, industry partner: Tomtom. Jan 2013 – March 2013.
  • “Early Detection of Road Traffic Incidents using Social Media”, industry partner: the Transport Management Centre; Oct – Dec 2012.
  • “Causal Inference for Sequential Traffic Congestion", industry partner: Microsoft Research Asia; Nov 2010 – Mar 2011.
  • “Abnormal Claim Detection from Worker’s Compensations”, industry partner: CGU Insurance; Mar 2010 – Jun 2011.
  • “Data Integration for Cross-Market Capital Trading Systems”, industry partner: the SMARST Group (now purchased by Nasdaq), Jun 2008 – Dec 2009.

Image of Wei Liu
Lecturer, A/DRsch Advanced Analytics Institute
Core Member, Advanced Analytics Institute
Member, Institute of Electrical and Electronics Engineers
Member, Association for Computing Machinery
+61 2 9514 3782

Research Interests

Main Research Interests:

  • Graph mining, dynamic network analysis, tensor factorization
  • Causal inference, Granger causality
  • Game theoretical modeling, adversarial learning 
  • Data imbalance, cost-sensitive learning
  • Anomaly (outlier) detection
Can supervise: Yes

Competitive PhD scholarships are available for prospective local and international research students.

Data Mining and Knowledge Discovery; Data Analytics.


Jiang, X., Liu, W., Cao, L. & Long, G. 2015, 'Coupled Collaborative Filtering for Context-aware Recommendation', AAAI Publications, Twenty-Ninth AAAI Conference on Artificial Intelligence, Student Abstracts, AAAI 2015, AAAI, Austin Texas, USA, pp. 4172-4173.
View/Download from: UTS OPUS
Context-aware features have been widely recognized as important factors in recommender systems. However, as a major technique in recommender systems, traditional Collaborative Filtering (CF) does not provide a straight-forward way of integrating the context-aware information into personal recommendation. We propose a Coupled Collaborative Filtering (CCF) model to measure the contextual information and use it to improve recommendations. In the proposed approach, coupled similarity computation is designed to be calculated by interitem, intra-context and inter-context interactions among item, user and context-ware factors. Experiments based on different types of CF models demonstrate the effectiveness of our design.