Belief Data for Recommendation Systems

Greetings from the weekly AI, ML and Friends Seminar at the Australian National University School of Computing, where Prof. Guy Aridor from Northwestern Kellogg School of Management is speaking on "The Value of Belief Data in Online Recommendation Systems". These are the systems which online stores use to recommend a product or post, based on what your friends like. Prof. Aridor's approach is to extract what each person likes. Such systems are worth billions of dollars and can also influence the outcome of elections.