Hybrid Framework Model for Group Recommendation

Group-based recommender systems are known for web-based applications to satisfy the preferences of every user in the group equally. The recommender system aims to identify user-preferred items, such as movies, music, books, and restaurants that tend to satisfy each individual and their collective needs in a group. They have a proclivity to solve the problem of information overload by probing through large volumes of dynamically generated information and provide its users with significant content and services. For better and more pertinent recommendations for a group of users, it is a non-trivial task as there is a huge diversity in tastes and preferences among various members of the group residing at different locations. This paper presents the results computed from traditional filtering approaches and the Hybrid filtering mode. These approaches are evaluated using various offline metrics on Movie-Lens dataset and the proposed model is based on the working efficiency and the quality of recommendations generated. This Hybrid Filtering Model proposed as Hyflix is explored for generating recommendations to a group of users and compared with the existing traditional filtering approaches in a social network.