EigenRec: Generalizing PureSVD for Effective and Efficient Top-N Recommendations

Abstract

We introduce EigenRec; a versatile and efficient Latent-Factor framework for Top-N Recommendations that includes the well-known PureSVD algorithm as a special case. EigenRec builds a low dimensional model of an inter-item proximity matrix that combines a similarity component, with a scaling operator, designed to control the influence of the prior item popularity on the final model. Seeing PureSVD within our framework provides intuition about its inner workings, exposes its inherent limitations, and also, paves the path towards painlessly improving its recommendation performance. A comprehensive set of experiments on the MovieLens and the Yahoo datasets based on widely applied performance metrics, indicate that EigenRec outperforms several state-of-the-art algorithms, in terms of Standard and Long-Tail recommendation accuracy, exhibiting low susceptibility to sparsity, even in its most extreme manifestations -- the Cold-Start problems. At the same time EigenRec has an attractive computational profile and it can apply readily in large-scale recommendation settings.

23 pages. Journal version of the conference paper "Factored Proximity Models for Top-N Recommendation"

References in corpus (2)

• Feature-based factorized Bilinear Similarity Model for Cold-Start Top-n Item Recommendation
• Cucheb: A GPU implementation of the filtered Lanczos procedure

Cited by in corpus (5)

• HybridSVD: When Collaborative Information is Not Enough
• The complementarity of a diverse range of deep learning features extracted from video content for video recommendation
• Boosting Item-based Collaborative Filtering via Nearly Uncoupled Random Walks
• Recommending on graphs: a comprehensive review from a data perspective
• Dynamic Modeling of User Preferences for Stable Recommendations