Graph Based Hybrid Approach for Long-Tail Item Recommendation in Collaborative Filtering

Recommender system plays a vital role in e-commerce business by providing personalized product recommendation. However, most of the existing recommender system are accuracy-centric and biases on recommending popular items. However, recommending relevant long tail items is another research challenge in recommendation community. In this article, we propose a graph based approach to enhance the long tail items in recommendation. The preliminary results are very encouraging on standard rating dataset. Proposed approach outperforms recently introduced recommender systems which focus on long tail item recommendation.