CIIR Talk Series: Maarten de Rijke

Speaker: Maarten de Rijke, University of Amsterdam

Title: The Power of Repeat

Date: Friday, December 9, 2022 - 1:30 - 2:30 PM EST (North American Eastern Standard Time) via Zoom. On campus attendees will gather in CS 151 to view the presentation. Refreshments provided.

Abstract: People are creatures of habit. We listen to the same song more than once and buy the same items regularly. Attention for repeat recommendations has been increasing steadily, due to the fact that user-item interactions show a lot of repetition in real-world scenarios. Several tasks related to repeat behavior have been studied, e.g., predicting whether the next purchase will be a consumed one, predicting which consumed item a user would like to buy again, and modeling item repeat inter-arrival times given explicit timestamps in the training data. 

In the talk I will report on recent progress in the development of recommendation algorithms for domains where repetition is a natural characteristic of human behavior. I will start with a bake-off between alternative approaches to next basket recommendation. I will then present new state-of-the-art methods for addressing the recommendation problem, consider alternative recommendation scenarios, and demonstrate the benefit of modeling and predicting repeat behavior in those scenarios.

Based on joint work with Mozhdeh Ariannezhad, Sami Jullien, Min Fang, Ming Li, Sebastian Schelter, and Andrew Yates.

Bio: Maarten de Rijke is a Distinguished University Professor of Artificial Intelligence and Information Retrieval at the University of Amsterdam. He is also the scientific director of the national Innovation Center for Artificial Intelligence (ICAI). His research is focused on designing trustworthy technology to connect people to information, particularly search engines, recommender systems, and conversational assistants. His work targets two key questions: How can we create intrinsic trust in information retrieval systems, that is, align their reasoning process with human expectations? And how can we establish extrinsic trust in information retrieval systems, that is, establish verifiable guarantees on their behavior?