GHRS: Graph-based Hybrid Recommendation System with Application to Movie\n Recommendation
No TL;DR found
Abstract
Research about recommender systems emerges over the last decade and comprises\nvaluable services to increase different companies' revenue. Several approaches\nexist in handling paper recommender systems. While most existing recommender\nsystems rely either on a content-based approach or a collaborative approach,\nthere are hybrid approaches that can improve recommendation accuracy using a\ncombination of both approaches. Even though many algorithms are proposed using\nsuch methods, it is still necessary for further improvement. In this paper, we\npropose a recommender system method using a graph-based model associated with\nthe similarity of users' ratings, in combination with users' demographic and\nlocation information. By utilizing the advantages of Autoencoder feature\nextraction, we extract new features based on all combined attributes. Using the\nnew set of features for clustering users, our proposed approach (GHRS) has\ngained a significant improvement, which dominates other methods' performance in\nthe cold-start problem. The experimental results on the MovieLens dataset show\nthat the proposed algorithm outperforms many existing recommendation algorithms\non recommendation accuracy.\n