Wide & Deep Learning for Recommender Systems

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.

Cited by in corpus (31)

• KGAT: Knowledge Graph Attention Network for Recommendation
• Neural Collaborative Filtering
• DKN: Deep Knowledge-Aware Network for News Recommendation
• Neural Factorization Machines for Sparse Predictive Analytics
• Meta-Prod2Vec - Product Embeddings Using Side-Information for Recommendation
• Attentional Factorization Machines: Learning the Weight of Feature Interactions via Attention Networks
• ATRank: An Attention-Based User Behavior Modeling Framework for Recommendation
• TF.Learn: TensorFlow's High-level Module for Distributed Machine Learning
• Time Series Classification from Scratch with Deep Neural Networks: A Strong Baseline
• Deep & Cross Network for Ad Click Predictions
• Deep Session Interest Network for Click-Through Rate Prediction
• Behavior Sequence Transformer for E-commerce Recommendation in Alibaba
• Neural News Recommendation with Attentive Multi-View Learning
• Pixie: A System for Recommending 3+ Billion Items to 200+ Million Users in Real-Time
• Real-Time Open-Domain Question Answering with Dense-Sparse Phrase Index
• PBODL : Parallel Bayesian Online Deep Learning for Click-Through Rate Prediction in Tencent Advertising System
• Sequential Scenario-Specific Meta Learner for Online Recommendation
• Large-scale Collaborative Filtering with Product Embeddings
• Related Pins at Pinterest: The Evolution of a Real-World Recommender System
• Science Driven Innovations Powering Mobile Product: Cloud AI vs. Device AI Solutions on Smart Device
• Joint Neural Collaborative Filtering for Recommender Systems
• DeepRec: An Open-source Toolkit for Deep Learning based Recommendation
• Getting deep recommenders fit: Bloom embeddings for sparse binary input/output networks
• Specializing Joint Representations for the task of Product Recommendation
• Session-based Sequential Skip Prediction via Recurrent Neural Networks
• Operation-aware Neural Networks for User Response Prediction
• Deep Bayesian Multi-Target Learning for Recommender Systems
• Res-embedding for Deep Learning Based Click-Through Rate Prediction Modeling
• Iterative Multi-document Neural Attention for Multiple Answer Prediction
• Columnar Database Techniques for Creating AI Features
• Field-aware Neural Factorization Machine for Click-Through Rate Prediction