Keynotes

Keynotes

Dr. Xiaoli Li, Data analytics department head at the Institute for Infocomm Research, A*STAR, Singapore

“Machine learning for Recommendation Systems”

Abstract: In recent years, personalized recommendation system has attracted a lot of research attention, and many methods have been developed to solve the recommendation problems under different application scenarios. In this talk, I will first present some background about personalized recommendation systems. Then, I will introduce representative machine learning techniques, e.g., learning-to-rank and deep learning, which have been used to build recommendation models. Finally, I will conclude my talk by discussing potential directions for future research in this area.


Dr. Xia Hu, Assistant Professor at Texas A&M University, USA

“Recommender Systems in “DEEP”: Architectures and Insights”

Abstract: This talk will cover some recent progress of recommender systems, from “DEEP” perspectives in both architecture and insight. Comparing with the conventional approaches, “DEEP” conception plays an important role in performance enhancement and result explanation. Specifically, deep architectures have been developed to improve  recommendation performance and deep insights are used to interpret the recommendation results. This talk will introduce how to integrate the “DEEP” aspects into recommender systems. First, as for the deep architecture, we strive to develop techniques based on neural networks to tackle the key problem in recommendation, i.e., collaborative filtering, on the basis of implicit feedback. By replacing the inner product with a neural architecture that can learn an arbitrary function from data, we show a general framework named NCF, short for Neural network based Collaborative Filtering. Second, as for the deep insight, we propose a post-hoc method called Sorted Explanation Paths (SEP) to interpret recommendation results. With appropriate metrics designing and explanation sorting, top-ranked explanation paths are selected as the final interpretation for deep insights into the targeting recommender systems. Third, stepping from the deep architecture and insight, we will also introduce some promising directions on the basis of the “DEEP” conception, aiming to inspire more researchers on the further development of this field.

Leave a Reply

Your email address will not be published. Required fields are marked *