Group and Single-user Influence Modeling for Personalized Recommendation
Michal Kompan · 2014
There are plenty of activities performed in the group instead of the single user. This can be observed in the digital world respectively. Various activities are performed over the Web or at least are discussed and agreed within the Web, e.g., social networks. For such scenarios the group recommendation is needed. Our work is focused on the improvement of recommendation approaches for group of users by introducing the inter group processes and members characteristics. Next, we proposed improvements in single-user recommendations by enhancing it with user context or virtual communities, which we treat similarly as groups in group recommenders. Proposed approaches are evaluated in several oine experiments, where the standard datasets are used. In order to investigate group-based features of proposed approaches, we performed user studies experiments as well.