Building Movie Recommender Systems Utilizing Poster's Visual Features: A Survey Study

Information overload has made it difficult for users to get needed data. In this context, recommender systems are used to filter information and provide personalized recommendations to help individuals make better choices. Providing consumers with a customized selection of movies to watch is critical for media service providers. However, sparsity and cold-start are two barriers to reliable recommendations. Recent advancements in deep learning architectures have spurred several studies to overcome recommender flaws using various informative resources, such as movie posters. In contrast to other surveys on movie recommendations, this work provides a comprehensive review of how visual data, especially movie posters, can enhance the results. Furthermore, comparing different approaches for building recommender systems is followed by in-depth surveying about investigating the movie recommender systems, including their methods, evaluation metrics, and datasets. This exhaustive analysis provides a detailed picture of the topic's popularity, gaps, and unexplored regions. It is envisaged that the proposed research and introduced possible future directions would serve as a stepping stone for researchers interested in building approvingly efficient modern movie recommender systems.