Dr Huizhi Liang, Department of Computer Science, University of Reading
Representation Learning for Recommender Systems
Representation Learning is to learn how to identify and disentangle the underlying explanatory factors hidden in the observed milieu of low-level sensory data. It is the fundamental goal of an artificial intelligence. The complex, huge, diverse, and dynamic data drives the urgent need for novel representation learning techniques. In this talk, I will discuss representation learning approaches such as deep neural networks based on user rating behavior data, social tagging behavior data, and social media data to discover knowledge about human users such as their interests, sentiment orientations, and information needs. The learned representations were applied in recommender systems to generate personalized recommendations.
Dr. Huizhi (Elly) Liang is a lecturer of the Department of Computer Science of the University of Reading. Her research interests include recommender systems, data mining, and machine learning. She did her PhD study in Queensland University of Technology, Australia in 2011. Before joining University of Reading, she worked as a postdoctoral research at Australian National University, The University of Melbourne, and LIP6, Sorbonne University and French National Centre for Scientific Research (CNRS).