Performance Improvement of Movie Recommender System Using Spectral Bi-clustering with Mahalabonis Distance

Collaborative Filtering with clustering is the primary method of recommendation. This research work introduced a new Spectral Bi-clustering with the Mahalabonis distance-based Movie Recommendation algorithm for Collaborative Filtering. In this paper, we compare the performance of several clustering methods like Kmeans, Spectral clustering with radial Basis Function and nearest neighbors affinity, Spectral clustering with radial Basis Function and nearest neighbors affinity with Mahalabonis distance, Spectral Bi-clustering with Mahalabonis distance to generate a movie recommendation system. Our experimental results show a significant performance improvement of our proposed Spectral Bi-clustering with the Mahalanobis distance-based Movie Recommendation algorithm over the traditional K-means algorithm as well as the Spectral Clustering algorithm with nearest neighbors and radial Basis Function affinity for Movie Recommender System.