![Loading Events](/wp-content/plugins/the-events-calendar/src/resources/images/tribe-loading.gif)
Guest Lecture: Assoc. Prof. Chirag Shah “Task-Based Intelligent Retrieval and Recommendation”
18.3.2022 | 14:00–15:30
Assoc. Prof. Chirag Shah from University of Washington is visiting us next week, and he will give a talk on his recent research concerning task-based search and recommendation systems. Welcome!
Time and place: Friday 18th of March 2022, 14 o’clock in Exactum, CK112 at Kumpula campus of the University of Helsinki (Pietari Kalmin katu 5, Helsinki)
Title: Task-Based Intelligent Retrieval and Recommendation
Abstract:
A majority of work in information retrieval (IR) and recommender systems assumes the existence of a well-defined query, a question, or at least an abstract sense of a need to be able to provide recommendations to a user. However, many situations and applications today pose a cold-start problem — needing to provide recommendations without having a clear query or sufficient data from the past. Some other applications could benefit by capturing and incorporating user’s task information for recommending related queries and item. In this talk, I will present a new research direction called Information Fostering, which involves identifying problems or opportunities in an information interaction episode without relying on a user’s expressed need — e.g., in cases of incomplete or vague queries. Addressing this problem requires going deeper into what, why, and how the user is engaged in that information interaction. Specifically, I will talk about some of our works to create a comprehensive task model with behavioral data, and using it for personalization and recommendation applications. The results show that considering task information creates more useful and more relevant recommendations, but more importantly, this approach can help address situations of zero-query and zero-intent, while also providing new user experiences in many applications, including e-commerce, productivity, and conversational agents.
Short Bio:
Chirag Shah is an Associate Professor in Information School (iSchool) at University of Washington (UW) in Seattle. He is also an Adjunct Associate Professor with Paul G. Allen School of Computer Science & Engineering (CSE) as well as Human Centered Design & Engineering (HCDE). Dr. Shah is the founding Director of InfoSeeking Lab and the Founding Co-Director of RAISE, a Center for Responsible AI. Before UW, he was at Rutgers University. He also holds posts of Visiting Professor at Peking University in China and Research Associate at University of Pretoria in South Africa. His research interests include intelligent search and recommender systems, trying to understand the task a person is doing and providing proactive recommendations. In addition to creating task-based IR systems that provide more personalized reactive and proactive recommendations, he is also focusing on making such systems transparent, fair, and free of biases. His research has resulted in 150 peer-reviewed papers and 6 books. He is the 2019 winner of British Computing Society’s Karen Spärck Jones award for significant contributions to IR and NLP. He has spent sabbaticals and summers visiting places like Spotify, Amazon, MSR AI, and Getty Images to work on problems involving search, recommendation and personalization, as well as designing novel customer experiences. He enjoys mindful walks, biking, painting, and ballroom dancing (when not socially distancing).
If you would like to meet and talk to him, please contact Giulio Jacucci (UH) and/or Thanh Tung Vuong (UH) so it can be arranged. There should be (some) time available for that after the talk as well.