An Embedding-based Deep Learning Approach for Movie Recommendation

Abstract

Recommender systems are employed either as tools or algorithms, whose key task is to efficiently predict the ratings for items and recommend items, using the data generated by users. It assists users in finding items that they will like. Hence, recommendation systems are becoming an essential part of some applications, e-commerce websites, and online streaming services, etc. This paper emphasizes on recommendation system for movies whose main objective is to propose a movie recommendation system through a deep learning technique. The size and complication of websites have increased due to the rapid growth of the internet. On these websites, it has become time-consuming and extremely difficult for the users to find the information that they are searching for. Therefore, a collaborative filtering-based movie recommendation system using deep learning and embedding is proposed. The proposed system is evaluated by calculating the RMSE and MAE values. The proposed method is compared with other machine learning algorithms and some state-of-the-art methods. Our model gave an MAE score of 0.7305 and an RMSE score of 0.9311. It performs better than some previous approaches to collaborative-filtering based movie recommendation.