A SYSTEMATIC REVIEW OF MOVIE RECOMMENDER SYSTEMS

Abstract

Recommender systems are vital to everyone's information selection. Managing massive amounts of data is common with recommendation system technology. Annual film releases are rising, and currently films are released within months. With movie releases, apps like Netflix, Viu, Amazon Prime Video, Disney+, etc. have emerged. Thus, Movie Recommender Systems (MRS) are essential to simplify and improve user experience. This research gives a systematic literature review (SLR) of MRS's current condition. Our comprehensive review addresses recommendation algorithms, data processing, and evaluation approaches. In SLR MRS, content-based filtering, collaborative filtering, knowledge-based recommender systems, and hybrid approaches are employed. To achieve this, 66 high-quality studies were selected from 27,187 2019-2023 studies using strict quality criteria. The study found that most MRSs use content-based filtering and machine learning to deliver non-personalized movie suggestions in various domains. The review helps researchers choose MRS development strategies. This study can assist MRS development catch up to other recommendation systems by improving efficiency. The MRS investigation found accuracy, sparsity, scalability, cold start, and operating time issues. Future study will examine how temporal and demographic data affect movie recommendation system relevancy and customization.