Visiting Lecturer Discusses Recommender Systems

James Caverlee, associate professor in the Department of Computer Science and Engineering at Texas A&M University, will give a seminar at 12:45 p.m. Thursday, Jan. 17, in the J.B. Hunt Transport Services Center, room 239. The seminar is titled "Recommender Systems: Fairness, Manipulation and Influentials."

Recommender systems are essential conduits for how we engage with the world: they shape the media we consume, the products we buy, the jobs we seek, and the friendships and professional contacts that form our social circles. In this talk, Caverlee will highlight three fundamental challenges to the ongoing success of recommenders, as well as some initial thoughts on how to overcome these challenges:

1. Fairness: since recommenders may be subject to algorithmic bias that can lead to negative consequences in the kinds of recommendations that are made (e.g., exhibiting gender bias or political bias), how can we design new fairness-aware algorithms that can empower users by enhancing diversity of topics and viewpoints?
2. Manipulation: since recommenders may be targeted by sophisticated campaigns of manipulation, how can we improve the robustness of these systems to such attacks? 
3. Influentials: many platforms allow the bottom-up discovery of influencers — users who provide unique specialized expertise, trustworthiness in decision-making, and access to novel content. How can we incorporate these influentials into recommenders while balancing the needs of fairness and robustness to manipulation?

Caverlee's research targets topics from social media, information retrieval, recommender systems, data mining, and emerging networked information systems. Caverlee is a recipient of a Young Faculty Award from the Defense Advanced Research Projects Agency, a Young Investigator Award from the Air Force Office of Scientific Research, an award from the National Science Foundation's Faculty Early Career Development Program, and several Google Research awards. He spent his sabbatical in 2015 at Google as a Visiting Scientist. Caverlee received his doctorate from Georgia Tech in 2007, masters degrees in computer science (2001) and in engineering-economic systems and operations research (2000) from Stanford University, and a bachelor's degree in economics from Duke University (1996).