Abstract:Matrix approximation is one of the most effective methods for collaborative filtering-based recommender systems. However, the high computation complexity of matrix factorization on large datasets limits its scalability. Prior solutions have adopted co-clustering methods to partition a large matrix into a set of smaller submatrices, which can then be processed in parallel to improve scalability. The drawback is that the recommendation accuracy is lower as the submatrices only contain subsets of the user-item rating information. This paper presents WEMAREC, a weighted and ensemble matrix approximation method for accurate and scalable recommendation. It builds upon the intuition that (sub)matrices containing more frequent samples of certain user/item/rating tend to make more reliable rating predictions for these specific user/item/rating. WEMAREC consists of two important components: (1) a weighting strategy that is computed based on the rating distribution in each submatrix and applied to approximate a single matrix containing those submatrices; and (2) an ensemble strategy that leverages user-specific and item-specific rating distributions to combine the approximation matrices of multiple sets of co-clustering results. Evaluations using real-world datasets demonstrate that WEMAREC outperforms state-of-the-art matrix approximation methods in recommendation accuracy (0.5?11.9% on the MovieLens dataset and 2.2--13.1% on the Netflix dataset) with 3--10X improvement on scalability.

Matrix Approximation with Cumulative Penalty for Collaborative Filtering

Mixture Matrix Approximation for Collaborative Filtering

MPMA: Mixture Probabilistic Matrix Approximation for Collaborative Filtering

Modeling user and item biases with Gaussian distribution for collaborative filtering

Wemarec: Accurate and Scalable Recommendation Through Weighted and Ensemble Matrix Approximation

Information Filtering Via Collaborative User Clustering Modeling.

GLOMA: Embedding Global Information in Local Matrix Approximation Models for Collaborative Filtering.

Synergizing Local and Global Models for Matrix Approximation

Preliminary data-based matrix factorization approach for recommendation

Ermma: Expected Risk Minimization for Matrix Approximation-Based Recommender Systems

AdaError

A two-step Recommendation Algorithm via Iterative Local Least Squares

Deep Matrix Factorization for Recommender Systems with Missing Data Not at Random

Multi-Task Matrix Factorization For Collaborative Filtering

Incorporating Community Detection and Clustering Techniques into Collaborative Filtering Model

Matrix Factorization Meets Cosine Similarity: Addressing Sparsity Problem In Collaborative Filtering Recommender System

Accelerated singular value thresholding for matrix completion.

Collaborative Filtering with Stability.

Collaborative Filtering with Noisy Ratings

Integrating user-side information into matrix factorization to address data sparsity of collaborative filtering

Item Similarity Learning Methods for Collaborative Filtering Recommender Systems