Market-Based Recommendations: Design, Simulation and Evaluation

This paper reports on the design, implementation, and evaluation of a market-based recommender system that suggests relevant documents to users. The key feature of the system is the use of market mechanisms to shortlist recommendations in decreasing order of user perceived quality. Essentially, the marketplace gives recommending agents the incentive to adjust their bids to different levels according to their belief about the corresponding user perceived quality. In order to test the efficiency of our marketplace design, this paper reports on our simulation results for different types of users with different information needs. In this context, we demonstrate that the bids from recommendations with different user perceived quality levels converge at different price levels and that the bidding agents can relate their bids to their internal belief about the quality of their recommendations. Keywords Recommender System, Auctions, Market-Based Approach

• Type: Conference
• Year: 2003
• Where: AOIS
• Authors: Yan Zheng Wei, Luc Moreau, Nicholas R. Jennings
• Keywords: AOIS 2003, Intelligent Agents, Market-based Recommender, Recommender System, User Perceived Quality