Adaptive Selection: An Analysis of Critiquing and Preference-Based Feedback in Conversational Recommender Systems
Lorraine McGinty and Barry Smyth
International Journal of Electronic Commerce,
Volume 11, Number 2, Winter 2006-07, pp. 35.
Abstract: E-commerce recommender systems help consumers to locate products within a complex product-space. Conversational recommender systems engage the user in a multi-cycle session, suggesting one or more products during each cycle, and using the feedback to inform the suggestions for the next cycle. By combining user feedback over several cycles, the system obtains a clear picture of the product the user wishes to purchase. As demonstrated under several experimental conditions, the performance of recommender systems is dramatically improved by the technique of adaptive selection, which employs critiquing and preference-based feedback, and emphasizes product diversity rather than similarity as a selection constraint.
Key Words and Phrases: Adaptive e-commerce applications, conversational recommender systems, critiquing, feedback elicitation, preference-based feedback, similarity and diversity.