Year: 2025
Type: conference-paper
Abstract: Relevance assessment is fundamental to Interactive Information Retrieval (IIR), and understanding the underlying cognitive processes is essential for optimizing user experience.To investigate how relevance and confirmation bias relate to cognitive load, we conducted a within-subject, lab-based, eye-tracking experiment (N=32).We applied the Low/High Index of Pupillary Activity (LHIPA) to measure cognitive load.LHIPA was calculated based on changes in pupil diameter and was unaffected by lighting conditions.Our results suggest that during relevance assessment: (1) for topically irrelevant documents, cognitive load is higher when users perceive them as relevant; for topically relevant documents, cognitive load is higher when users perceive them as irrelevant; (2) users with higher levels of topic familiarity and interest expend more cognitive load; and(3) users with higher predisposition to confirmation bias invest less cognitive load.These findings demonstrate that during relevance assessment, users' perceived relevance, topical knowledge, interest, and confirmation bias tendency influence cognitive load levels.This research has implications for improving user interface design to facilitate more accurate relevance judgments and more effective, unbiased information retrieval.
Authors: Li Shi, Gavindya Jayawardena, Jacek Gwizdka
Institution: The University of Texas at Austin
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
Funder: University of Texas at Austin
OpenAlex ID: W4409973460