The Center for Intelligent Information Retrieval (CIIR) hosted a Brainstorming Session on “Information Retrieval (IR) Research in the Age of Generative Artificial Intelligence (AI)” that was held on December 1 – 2, 2023 at UMass Amherst. CIIR faculty organizers, James Allan, Negin Rahimi, and Hamed Zamani, invited a small group of experts to this two-day workshop to delve into the myriad challenges and opportunities presented by the integration of generative AI and IR.
“In today's rapidly evolving digital landscape, the field of information retrieval stands at the intersection of traditional search methodologies and cutting-edge machine learning and artificial intelligence technologies,” said Zamani. “As we witness the proliferation of generative AI-driven models, such as diffusion and large language models, it becomes increasingly evident that the boundaries of IR research are expanding.” To explore and navigate this dynamic landscape, the brainstorming session was organized with the goal of charting the course for the future of IR research, where generative AI plays a central role in reshaping how we discover and interact with information.
The workshop included presentations covering topics from using search to build more accurate and more useful language models to generative approaches to IR itself. Discussions covered the presentations as well as personalization, recommendation, conversational search, retrieval augmented LMs, attribution, and trust. Workshop presenters were: James Allan (UMass Amherst), Samuel Carton (U. of New Hampshire), Bruce Croft (UMass Amherst), Laura Dietz (U. of New Hampshire), Mohit Iyyer (UMass Amherst), Vanessa Murdock (Amazon), Jeremy Pickens (Redgrave Data), Negin Rahimi (UMass Amherst), Chirag Shah (U. of Washington), Hamed Zamani (UMass Amherst), and Yongfeng Zhang (Rutgers U.). A number of faculty members and students from the CIIR and the broader Manning College of Information and Computer Sciences were also present and participated in the discussions.
The outcome of the event will help the CIIR organizers to plan a larger workshop soon with a broader audience, with the goal of developing a broadly agreed-upon vision and plan that integrates generative AI and Information Retrieval.
Photo: Faculty, industry, and student attendees at the IR Brainstorming Session.