Realization of Recommender Framework Based on Community Detection
Recommender systems play an important role in suggesting relevant information to users based on their available preferences about items. Utilizing a recommender system allows companies to increase revenues, customer satisfaction and enable personalization and discovery. Content-based and collaborative filtering approaches are the most popular techniques in recommender systems predicting users preferences based on “collaborative” data about users and items in the system. However, their use is not justified in certain applications, particularly when user-item collaboration data is very sparse or missing. In this paper, a recommender framework based on community detection is developed outperforming other popular recommendation methods in some applications.