EigenRec: Generalizing PureSVD for Effective and Efficient Top-N Recommendations
This work addresses the need for more effective and scalable recommendation systems, particularly in handling sparsity and cold-start problems, though it is incremental as it builds upon and improves the existing PureSVD method.
The paper tackles the problem of improving Top-N recommendation accuracy and efficiency by introducing EigenRec, a latent-factor framework that generalizes PureSVD, and it shows that EigenRec outperforms state-of-the-art algorithms on MovieLens and Yahoo datasets in terms of standard and long-tail accuracy, with low susceptibility to sparsity and cold-start issues.
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.