Graph Representation Learning for Recommendation Systems: A Short Review

Since the information explosion, a large number of items are present on the web, making it difficult for users to find the appropriate item from the available set of options. The Recommender System (RS) solves the problem of information overload by suggesting items of interest to the user. It has grown in popularity over the last few decades, and a significant amount of research has been conducted in this field. Among them, Collaborative Filtering (CF) is the most popular and widely used approach for RS, attempting to analyze the user’s interest in the target item based on the opinions of other like-minded users. But recent years have witnessed the fast development of the emerging topic of Heterogeneous information networks Recommender Systems. Heterogeneous Information Network (HIN) based recommender systems offer a unified approach to combining various additional information, which can be combined with mainstream recommendation algorithms to effectively improve model performance and interpretability, and have thus been applied in a wide range of recommendation tasks. This paper provides a brief overview of various approaches used for recommender systems, as well as an understanding of the Collaborative Filtering technique. We also discussed HIN-based techniques, and finally, we focus on research challenges that must be addressed.