Addressing the Cold Start Challenge through the Implementation of Content-based Filtering and Hybrid Filtering for Movie Recommendation
Mahi Kolli, L D Mukil, N Kartthikeyan, M Pranavkrishnan, Nalini Sampath, Priyanka C. Nair · 2024
Recommender systems are of great significance for the difficulty in increasing speed as well as the amount of online information of the users to sort out the relevant content as they offer personalized suggestions. Content-based filtering and collaborative filtering are the two very common types of recommender algorithms, but they are still facing some issues, e.g. sparsity and cold-start problems. To cope with these limitations and increase recommendation accuracy, this study proposes a new form of recommendation system that combines collaborative filtering and content-based filtering to form a hybrid recommender system. The system proposed integrates the advantages of both methods and, therefore, offers more precise and individualized recommendations. As an illustration, the hybrid method shows superiority compared to the traditional methods, which proves that deep learning might be the key to advancing recommendation systems.