Graph Neural Network based Recommender System

Abstract

With the rapid rise of online content, Recommender Systems are becoming increasingly important in reducing information overload. Because of the important technology worth of recommendation system, new studies on the subject have been published on a regular basis. The major difficulty in recommender systems is to create suitable user/item models using user activities as well as other data. Graph neural network techniques have lately become widely employed in recommender systems because most of the data is fundamentally graph architecture, and Graph Neural Network has advantage in graph supervised learning. This Paper aims to provide recent work done in the Graph Neural Network based recommender system. The paper also discusses the open issues and challenges pertaining to Graph Neural Network.