Deep Heterogeneous Autoencoders for Collaborative Filtering

TL;DR

This paper introduces a deep autoencoder-based model that integrates heterogeneous auxiliary data to improve recommendation accuracy in sparse data environments, capturing diverse data types and user dynamics.

Contribution

It presents a novel autoencoder architecture for multiple data sources, enhancing collaborative filtering by effectively modeling heterogeneous information.

Findings

Improved mean average precision on MovieLens datasets.

Enhanced recall rates over existing methods.

Effective modeling of user preferences and item features.

Abstract

This paper leverages heterogeneous auxiliary information to address the data sparsity problem of recommender systems. We propose a model that learns a shared feature space from heterogeneous data, such as item descriptions, product tags and online purchase history, to obtain better predictions. Our model consists of autoencoders, not only for numerical and categorical data, but also for sequential data, which enables capturing user tastes, item characteristics and the recent dynamics of user preference. We learn the autoencoder architecture for each data source independently in order to better model their statistical properties. Our evaluation on two MovieLens datasets and an e-commerce dataset shows that mean average precision and recall improve over state-of-the-art methods.