Video content recommendation for group based on viewing history and viewer preference

This paper proposes an algorithm to estimate the useful content for known groups. The method makes it possible to recommend for the group such as friends, couple and family. As the first step, we focused on the preference between group members. Our algorithm estimates the preference using rating of individual for video genres and viewing history shared together. Then we judge that the content is useful for the group based on the preference. Evaluation tests show that our proposal algorithm is serendipitous.