Boosting Trust in Collaborative Recommender Agents with Interest Similarity

Abstract

Inserted in communities of people with similar interests, recommender agents predict the behavior of users based on the behavior of other like-minded people. In addition to user similarity, trustworthiness is a factor that agents have to consider in the selection of reliable partners for collaboration. Previous works focused on modeling trust in recommender systems base on global user profile similarity or history of exchanged opinions. In this paper we propose a novel approach for agent-based recommendation in which trust is independently learned and evolved for each pair of interest topics two users have in common. Experimental results show that agents learning who to trust about certain topics reach better levels of precision than considering exclusively user similarity.