A Collaborative Filtering Movies Recommendation System based on Graph Neural Network
Redwane Nesmaoui, Mouad Louhichi, Mohamed Lazaar
Abstract
The implementation of machine learning algorithms in marketing by organisations has been more beneficial in recent years. Overall, it has become a major contributor to a company's success and development in terms of growth and income since it helps to recommend the interesting product/service to the right individuals or groups without requiring them to go through a long complex procedure to receive an interesting item from a list of millions, in the other side Graph Neural Network is used widely in the recent machine learning applications including Recommender Systems. The purpose of this research is the evaluation of a LightGCN Movies Recommendation System, and its efficiency in modelling and building relationship between movies, by providing suggesting new/unknown items to the users that will like them, those recommendations will be based on representing Movies as a node and their ratings as edges of the graph, which will help to build a continuous representation of nodes and edges, this approach required the combination of a classification model to predict the existence of the relationship between movies and their features as genres, release year, etc, this approach will enable us to predict when no neighbourhoods information is known.