Employing Sparsity Removal Approach and Fuzzy C-Means Clustering Technique on a Movie Recommendation System
Noor Ifada, Eko Hadi Prasetyo, Mula'ab Mula'ab · 2018
Collaborative Filtering (CF) approach has been used in many recommendation systems. Despite its popularity, CF faces several challenges such as data sparsity and scalability. In this paper, we propose a novel clustered item-based CF to solve both problems. To overcome the sparsity issue of rating data, we propose a novel sparsity removal approach that employs the combination of rating and movie genre similarities. To overcome the scalability issue, we apply the use the Fuzzy C-Means clustering technique to create groups of movies. Evaluating the proposed method on a real-world movie dataset, we show that our proposed method produces a dense rating data, is scalable for high dimensional data, and improves the recommendation quality of the traditional item-based CF method.