Towards trustworthy recommender systems

Abstract

Recommender systems are among the most widely used applications of Artificial Intelligence (AI), helping websites and apps connect users with relevant items such as news articles and products that align with their interests. Thanks to advancements in AI, modern recommender systems have achieved remarkable accuracy. However, their design often lacks consideration for social responsibility, leading to potential negative impacts, such as the promotion of fake content or biased recommendations. These issues pose significant risks to society by spreading misinformation and reinforcing biases. In this talk, I will first introduce the fundamental concepts and framework of trustworthy recommender systems. I will then present a novel AI-powered responsible recommender system designed to combat fake news, along with a specialized framework aimed at delivering accurate and unbiased content recommendations. Additionally, I will discuss research on fair recommendations in the context of incomplete data. Finally, I will conclude by highlighting promising directions for future advancements in this dynamic field.

Presenter

Shoujin Wang is a Lecturer in Data Science at University of  Technology Sydney. He was named in Stanford's List of World's Top 2% Scientists  in 2023 and 2024. Shoujin obtained his PhD in Data Science from University of  Technology Sydney in 2019. His main research interests include data science,  machine learning, recommender systems and misinformation mitigation.  He  has published more than 80 research papers in these areas, most of which were  published at premier data science and AI conferences or journals, like NeurIPS,  KDD, The WebConf, SIGIR, AAAI, IJCAI, TKDE, TIST and CSUR. His research has  been broadly reported by national media outlets including SBS, ABC Science  Show, etc.  Shoujin has generally served as a PC member or a senior PC  member for over 10 premier international data science conferences including  KDD, The WebConf, AAAI, IJCAI, WSDM, CIKM and a reviewer for more than 10  prestigious journals including Machine Learning, IEEE TKDE, ACM TOIS, etc.  Shoujin is the recipient of various prestigious awards and honors, e.g., the  2021 DAAD AINet Fellowship, 2022 Club Melbourne Fellowship, 2022 DSAA  Next-generation Data Scientist Award, 2023 Royal Society of NSW Bicentennial  Early Career Research and Service Citations Award and 2024 Heidelberg Laureate Forum Young Researcher Fellowship.