Collaborative Filtering with Label Consistent Restricted Boltzmann Machine

TL;DR

This paper enhances RBM-based collaborative filtering by incorporating user demographics and item metadata, leading to improved recommendation accuracy comparable to current state-of-the-art models.

Contribution

It introduces a label consistent RBM that integrates supervision through metadata, significantly improving upon previous unsupervised RBM approaches.

Findings

Significant performance improvement over previous RBM-based methods

Achieves results comparable to state-of-the-art latent factor models

Demonstrates the effectiveness of supervised RBM in recommender systems

Abstract

The possibility of employing restricted Boltzmann machine (RBM) for collaborative filtering has been known for about a decade. However, there has been hardly any work on this topic since 2007. This work revisits the application of RBM in recommender systems. RBM based collaborative filtering only used the rating information; this is an unsupervised architecture. This work adds supervision by exploiting user demographic information and item metadata. A network is learned from the representation layer to the labels (metadata). The proposed label consistent RBM formulation improves significantly on the existing RBM based approach and yield results at par with the state-of-the-art latent factor based models.