Towards Understandable Personalized Recommendations: Hybrid Explanations

Martin Svrcek, Michal Kompan, Maria Bielikova

Nowadays, personalized recommendations are widely used and popular. There are a lot of systems in various fields, which use recommendations for different purposes. One of the basic problems is the distrust of users of recommended systems. Users often consider the recommendations as an intrusion of their privacy. Therefore, it is important to make recommendations transparent and understandable to users. To address these problems, we propose a novel hybrid method of personalized explanation of recommendations. Our method is independent of recommendation technique and combines basic explanation styles to provide the appropriate type of personalized explanation to each user. We conducted several online experiments in the news domain. Obtained results clearly show that the proposed personalized hybrid explanation approach improves the users’ attitude towards the recommender, moreover, we have observed the increase of recommendation precision.