MarCol: A market-based recommender system

Dan Melamed, Bracha Shapira, Yuval Elovici

IEEE Intelligent Systems 22 (3), 74-78, 2007

Collaborative information-filtering systems maintain user judgments on the relevance of data items.These systems recommend relevant information to other users on the basis of similarity between user profiles or recommended items. This market-based collaborative information-filtering system employs a pricing mechanism to motivate users to provide judgments. Results show that the model increases feedback and improves recommendation quality. MarCol uses Google as the underlying search engine. It stores all user logs, including queries and judgments.