Optimizing preferences within groups: A case study on travel recommendation

Abstract

This work describes a multiagent recommender system where agents work on behalf of members of a group of customers, trying to reach the best recommendation for the whole group. The goal is to model the group recommendation as a distributed constraint optimization problem, taking customer preferences into account and searching for the best solution. Experimental results show that this approach can be sucessfully applied to propose recommendations to a group of users. © 2008 Springer Berlin Heidelberg.