Group recommender systems as a voting problem

Abstract

Nowadays, technology allows for a better understanding of user needs through system design (recommender system) methodologies that position the individual at the center of all his actions. In this paper we start by reviewing the state of the art in both individual and group recommender systems technologies. On this ground we cluster the main characteristics of recommender systems with respect to the tasks they perform, the methods they employ and the issues they address. The other theoretical part we rely on is derived from social choice theory and voting. The main objective of this paper is to highlight the role of voting in group recommender systems, more precisely discussing several voting methods together with their characteristics. Our main contributions focus on: reviewing the state of the art literature related to voting in GRS, proposing an innovative and transparent voting mechanism and highlighting the current development of our music recommender system, GroupFun. © 2013 Springer-Verlag Berlin Heidelberg.