Web-Based Movie Recommendation System using Content-Based Filtering and KNN Algorithm

Improving the effectiveness of approaches has been a significant obstacle for recommended systems. It has also been crucial to address the trade-off between accuracy and calculation time when promoting goods via recommendation systems, as they must deliver accurate and real-time recommendations. This article discusses the development of a web-based system for recommending movies based on content-based filtering techniques. The system can recommend movies based on any combination of these two or more characteristics. This study has created a recommendation system based on genre preferences. To achieve this, content-based filtering based on genre connection is employed using cosine similarity. In order to improve the accuracy of this system, the KNN algorithm and the cosine similarity concept were implemented. The system employed a Movie Lens dataset, and the accuracy reached more than 87%. Moreover, this research successfully developed a user-friendly movie recommendation system (RS) with Python data analytic tools and the Django web framework. Moreover, this system can be accessed using a web browser and has more than 87% accuracy for giving movie recommendations based on some features or characteristics of the movie.