Co-Authoring the Self: A Human-AI Interface for Interest Reflection in Recommenders
This work addresses the need for more transparent and trustworthy recommender systems for users, though it is incremental as it builds on existing natural language-based user profiles.
The paper tackled the problem of improving transparency and user trust in recommender systems by introducing a human-AI collaborative profile that allows users to edit and reflect on personalized interest summaries. The result, from an eight-week deployment with 1775 users, showed persistent gaps between user-perceived and system-inferred interests, with the profile encouraging engagement and reflection.
Natural language-based user profiles in recommender systems have been explored for their interpretability and potential to help users scrutinize and refine their interests, thereby improving recommendation quality. Building on this foundation, we introduce a human-AI collaborative profile for a movie recommender system that presents editable personalized interest summaries of a user's movie history. Unlike static profiles, this design invites users to directly inspect, modify, and reflect on the system's inferences. In an eight-week online field deployment with 1775 active movie recommender users, we find persistent gaps between user-perceived and system-inferred interests, show how the profile encourages engagement and reflection, and identify design directions for leveraging imperfect AI-powered user profiles to stimulate more user intervention and build more transparent and trustworthy recommender experiences.