Wide & Deep Learning for Recommender Systems

TL;DR

This paper introduces Wide & Deep learning, a hybrid model combining linear and neural network components to improve recommender systems by balancing memorization and generalization, demonstrated on Google Play with significant performance gains.

Contribution

The paper presents a novel joint training approach for wide linear models and deep neural networks to enhance recommender system performance.

Findings

Wide & Deep outperforms wide-only and deep-only models in app acquisition.

The system was successfully deployed at Google Play with over one billion users.

Open-sourced implementation in TensorFlow is provided.

Abstract

Generalized linear models with nonlinear feature transformations are widely used for large-scale regression and classification problems with sparse inputs. Memorization of feature interactions through a wide set of cross-product feature transformations are effective and interpretable, while generalization requires more feature engineering effort. With less feature engineering, deep neural networks can generalize better to unseen feature combinations through low-dimensional dense embeddings learned for the sparse features. However, deep neural networks with embeddings can over-generalize and recommend less relevant items when the user-item interactions are sparse and high-rank. In this paper, we present Wide & Deep learning---jointly trained wide linear models and deep neural networks---to combine the benefits of memorization and generalization for recommender systems. We productionized and evaluated the system on Google Play, a commercial mobile app store with over one billion active users and over one million apps. Online experiment results show that Wide & Deep significantly increased app acquisitions compared with wide-only and deep-only models. We have also open-sourced our implementation in TensorFlow.