Data Science Talk Series 

Bounded-Rationality-Aware Vulnerability Evaluation (BRAVE) in Interactive and Generative Information Retrieval (In-person)

Abstract: Generative and interactive information retrieval (IR) systems reshape how people seek, judge, and use information, yet human decisions remain bounded by attention, heuristics, and contexts. This talk introduces BRAVE, a bounded-rationality-aware vulnerability evaluation approach that quantifies how human users and models become susceptible to cognitive traps during text retrieval, conversational search, and information evaluation. Combining behavioral studies with user simulations, BRAVE estimates individuals’ cognitive limits in both single-query and conversation-level interactions, measures bias risks of retrieval and language models, and assesses the potential of bounded-rationality-aware interventions. Case studies in health and learning domains show how mental models and system vulnerability vary across bias triggers, model prompting, and interface affordances. This talk will conclude with policy implications for trustworthy information interactions – making systems that not only answer well, but also better protect bounded users engaging in dynamic and manipulative information spaces.

Bio of the speaker: Dr. Jiqun Liu is an Associate Professor in Data Science (iSchool) and an Affiliated Faculty in Psychology at the University of Oklahoma. His research focuses on information retrieval (IR), machine learning, and human-AI interaction, with a particular emphasis on cognitive biases identification, human-centered evaluation on interactive and generative systems, and scalable bias mitigation strategies. Dr. Liu’s work bridges behavioral science and IR, and investigates how boundedly rational users interact with search engines, recommender systems, and large language models (LLMs) in information interaction and decision-making tasks. His research on bias-aware user modeling and system evaluation received grant support from National Science Foundation (NSF), Institute of Museum and Library Services (IMLS), and Microsoft, and has been published at premier venues, such as ACM SIGIR, CHIIR, JASIST, IP&M, TOIS, EMNLP, and TheWebConf. His work has also been introduced in a research monograph entitled A Behavioral Economics Approach to Interactive Information Retrieval: Understanding and Supporting Boundedly Rational Users by Springer Nature and presented through invited talks, tutorials, and workshops to both academic audiences and industry practitioners.