Combining decentralized collaborative filtering on PDAs with multi-touch tabletops for group recommendation

We have designed and implemented an approach based on decentralized, item-based collaborative filtering to recommend items on Personal Digital Assistants (PDAs). The system exchanges rating vectors among the mobile devices, computes local matrices of item similarity and utilizes them to generate recommendations. In addition, our approach incorporates recommendations for groups of users that are present at a public shared display in a collocated setting. This system can be extended with interaction on a multi-touch tabletop to improve group recommendations using critique-based recommender techniques.