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 (108)

• KGAT: Knowledge Graph Attention Network for Recommendation
• Tensor Comprehensions: Framework-Agnostic High-Performance Machine Learning Abstractions
• 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
• Bandana: Using Non-volatile Memory for Storing Deep Learning Models
• Deep Session Interest Network for Click-Through Rate Prediction
• Deep Interest Network for Click-Through Rate Prediction
• AutoFIS: Automatic Feature Interaction Selection in Factorization Models for Click-Through Rate Prediction
• Behavior Sequence Transformer for E-commerce Recommendation in Alibaba
• It Is Different When Items Are Older: Debiasing Recommendations When Selection Bias and User Preferences Are Dynamic
• OneFlow: Redesign the Distributed Deep Learning Framework from Scratch
• Neural News Recommendation with Attentive Multi-View Learning
• DeepFM: An End-to-End Wide & Deep Learning Framework for CTR Prediction
• Joint Optimization of Tree-based Index and Deep Model for Recommender Systems
• Pixie: A System for Recommending 3+ Billion Items to 200+ Million Users in Real-Time
• Perceive Your Users in Depth: Learning Universal User Representations from Multiple E-commerce Tasks
• EENMF: An End-to-End Neural Matching Framework for E-Commerce Sponsored Search
• Developing a Recommendation Benchmark for MLPerf Training and Inference
• Billion-scale Commodity Embedding for E-commerce Recommendation in Alibaba
• Product-based Neural Networks for User Response Prediction over Multi-field Categorical Data
• Learning from Multi-View Multi-Way Data via Structural Factorization Machines
• Learnable Embedding Sizes for Recommender Systems
• Embedding-based Recommender System for Job to Candidate Matching on Scale
• Real-Time Open-Domain Question Answering with Dense-Sparse Phrase Index
• TabGNN: Multiplex Graph Neural Network for Tabular Data Prediction
• NeuRec: On Nonlinear Transformation for Personalized Ranking
• Mixture of Virtual-Kernel Experts for Multi-Objective User Profile Modeling
• PBODL : Parallel Bayesian Online Deep Learning for Click-Through Rate Prediction in Tencent Advertising System
• Fast, Accurate, and Simple Models for Tabular Data via Augmented Distillation
• A Probabilistic Framework for Location Inference from Social Media
• Click-aware purchase prediction with push at the top
• Sequential Scenario-Specific Meta Learner for Online Recommendation
• DNN or k-NN: That is the Generalize vs. Memorize Question
• On the impact of selected modern deep-learning techniques to the performance and celerity of classification models in an experimental high-energy physics use case
• Personalizing Similar Product Recommendations in Fashion E-commerce
• Large-scale Collaborative Filtering with Product Embeddings
• Related Pins at Pinterest: The Evolution of a Real-World Recommender System
• Transfer Meets Hybrid: A Synthetic Approach for Cross-Domain Collaborative Filtering with Text
• Science Driven Innovations Powering Mobile Product: Cloud AI vs. Device AI Solutions on Smart Device
• Joint Neural Collaborative Filtering for Recommender Systems
• Image Matters: Visually modeling user behaviors using Advanced Model Server
• Deep density networks and uncertainty in recommender systems
• MARS: Memory Attention-Aware Recommender System
• Neural Variational Hybrid Collaborative Filtering
• DeepRec: An Open-source Toolkit for Deep Learning based Recommendation
• Collaborative Metric Learning Recommendation System: Application to Theatrical Movie Releases
• Automatic Generation of Chinese Short Product Titles for Mobile Display
• DADNN: Multi-Scene CTR Prediction via Domain-Aware Deep Neural Network
• Visualizing and Understanding Deep Neural Networks in CTR Prediction
• Getting deep recommenders fit: Bloom embeddings for sparse binary input/output networks
• Specializing Joint Representations for the task of Product Recommendation
• Strategy and Benchmark for Converting Deep Q-Networks to Event-Driven Spiking Neural Networks
• Enhancing Model Interpretability and Accuracy for Disease Progression Prediction via Phenotype-Based Patient Similarity Learning
• Session-based Sequential Skip Prediction via Recurrent Neural Networks
• GAN-based Recommendation with Positive-Unlabeled Sampling
• Capacity Control of ReLU Neural Networks by Basis-path Norm
• Deep Reinforcement Learning based Group Recommender System
• Pairwise Interactive Graph Attention Network for Context-Aware Recommendation
• DeepLight: Deep Lightweight Feature Interactions for Accelerating CTR Predictions in Ad Serving
• Operation-aware Neural Networks for User Response Prediction
• Deep Bayesian Multi-Target Learning for Recommender Systems
• Learning to Recommend via Meta Parameter Partition
• Denoising User-aware Memory Network for Recommendation
• Predicting Natural Hazards with Neuronal Networks
• Self-Attentive Neural Collaborative Filtering
• Research Progress of News Recommendation Methods
• T-RECS: A Simulation Tool to Study the Societal Impact of Recommender Systems
• Multi-Graph Convolution Collaborative Filtering
• Res-embedding for Deep Learning Based Click-Through Rate Prediction Modeling
• Iterative Boosting Deep Neural Networks for Predicting Click-Through Rate
• Feature Interaction based Neural Network for Click-Through Rate Prediction
• Large-scale Hybrid Approach for Predicting User Satisfaction with Conversational Agents
• KNN, An Underestimated Model for Regional Rainfall Forecasting
• Multi-Level Deep Cascade Trees for Conversion Rate Prediction in Recommendation System
• Accelerated learning from recommender systems using multi-armed bandit
• Learning to Embed Categorical Features without Embedding Tables for Recommendation
• Context-aware Heterogeneous Graph Attention Network for User Behavior Prediction in Local Consumer Service Platform
• Deep Sequence Learning with Auxiliary Information for Traffic Prediction
• RPC Considered Harmful: Fast Distributed Deep Learning on RDMA
• Explainable Recommendation Systems by Generalized Additive Models with Manifest and Latent Interactions
• Differential Geometric Retrieval of Deep Features
• UBER-GNN: A User-Based Embeddings Recommendation based on Graph Neural Networks
• Gear Training: A new way to implement high-performance model-parallel training
• Columnar Database Techniques for Creating AI Features
• Role Semantics for Better Models of Implicit Discourse Relations
• Iterative Multi-document Neural Attention for Multiple Answer Prediction
• C-3PO: Click-sequence-aware DeeP Neural Network (DNN)-based Pop-uPs RecOmmendation
• Collaborative Multi-modal deep learning for the personalized product retrieval in Facebook Marketplace
• Multimodal Topic Learning for Video Recommendation
• Personalized Bundle Recommendation in Online Games
• Field-aware Neural Factorization Machine for Click-Through Rate Prediction
• Action-conditional Sequence Modeling for Recommendation
• Towards Deep and Representation Learning for Talent Search at LinkedIn
• Dual Adversarial Variational Embedding for Robust Recommendation
• Personalized Neural Embeddings for Collaborative Filtering with Text
• Learning Theory and Algorithms for Revenue Management in Sponsored Search
• Low-Precision Hardware Architectures Meet Recommendation Model Inference at Scale
• Automatic Historical Feature Generation through Tree-based Method in Ads Prediction
• Neural Collaborative Ranking
• URIR: Recommendation algorithm of user RNN encoder and item encoder based on knowledge graph
• Supporting Massive DLRM Inference Through Software Defined Memory