Argumentative Explanations for Group Recommender Systems

Abstract

This paper presents an approach in providing explanations for a group setting. In group recommendation systems, the preferences and interests of all group members must be taken into account in order to generate recommendations that are acceptable to the group as a whole. We propose a new approach to providing explanations to groups based on the idea of having a graphical chassis forming the basis for argumentative scaffolding. In this way, the approach aims to help group members understand and accept the group recommendations by showing how they are aligned both through a collaborative filtering and content based system. The proposed approach is evaluated through a user study, in which participants were asked to evaluate the quality of explanations generated by the proposed approach in terms of transparency, effectiveness, trust and privacy. The results are promising but not yet conclusive due to the low number of participants. This paper also discusses the challenges and limitations of designing explanations for group recommendation systems. We conclude the proposed approach is a step towards addressing these challenges and improving the usability and acceptability of group recommendation systems. Overall, this research contributes to the field by providing a new approach for designing explanations that is effective in helping group members understand and accept the group recommendations.