Ask the GRU: Multi-Task Learning for Deep Text Recommendations

Abstract

In a variety of application domains the content to be recommended to users is associated with text. This includes research papers, movies with associated plot summaries, news articles, blog posts, etc. Recommendation approaches based on latent factor models can be extended naturally to leverage text by employing an explicit mapping from text to factors. This enables recommendations for new, unseen content, and may generalize better, since the factors for all items are produced by a compactly-parametrized model. Previous work has used topic models or averages of word embeddings for this mapping. In this paper we present a method leveraging deep recurrent neural networks to encode the text sequence into a latent vector, specifically gated recurrent units (GRUs) trained end-to-end on the collaborative filtering task. For the task of scientific paper recommendation, this yields models with significantly higher accuracy. In cold-start scenarios, we beat the previous state-of-the-art, all of which ignore word order. Performance is further improved by multi-task learning, where the text encoder network is trained for a combination of content recommendation and item metadata prediction. This regularizes the collaborative filtering model, ameliorating the problem of sparsity of the observed rating matrix.

8 pages

References in corpus (11)

• Adam: A Method for Stochastic Optimization
• Distributed Representations of Words and Phrases and their Compositionality
• Sequence to Sequence Learning with Neural Networks
• Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling
• Natural Language Processing (almost) from Scratch
• BPR: Bayesian Personalized Ranking from Implicit Feedback
• Neural Network Matrix Factorization
• Extraction of Salient Sentences from Labelled Documents
• Learning Distributed Representations from Reviews for Collaborative Filtering
• VBPR: Visual Bayesian Personalized Ranking from Implicit Feedback
• Joint Training of Deep Boltzmann Machines

Cited by in corpus (46)

• Graph Convolutional Neural Networks for Web-Scale Recommender Systems
• Deep Learning based Recommender System: A Survey and New Perspectives
• A Survey on Accuracy-oriented Neural Recommendation: From Collaborative Filtering to Information-rich Recommendation
• TransNets: Learning to Transform for Recommendation
• Contextual Hybrid Session-based News Recommendation with Recurrent Neural Networks
• Learning to Recommend with Multiple Cascading Behaviors
• News Session-Based Recommendations using Deep Neural Networks
• A Correlative Denoising Autoencoder to Model Social Influence for Top-N Recommender System
• A Deep Multimodal Approach for Cold-start Music Recommendation
• Deep Learning for Sequential Recommendation: Algorithms, Influential Factors, and Evaluations
• Automatic Rule Extraction from Long Short Term Memory Networks
• MetaBalance: Improving Multi-Task Recommendations via Adapting Gradient Magnitudes of Auxiliary Tasks
• Understanding and Improving Fairness-Accuracy Trade-offs in Multi-Task Learning
• TLSAN: Time-aware Long- and Short-term Attention Network for Next-item Recommendation
• Quantitative analysis of Matthew effect and sparsity problem of recommender systems
• Joint Text Embedding for Personalized Content-based Recommendation
• MatRec: Matrix Factorization for Highly Skewed Dataset
• CATA++: A Collaborative Dual Attentive Autoencoder Method for Recommending Scientific Articles
• Transfer Meets Hybrid: A Synthetic Approach for Cross-Domain Collaborative Filtering with Text
• Joint Neural Collaborative Filtering for Recommender Systems
• Flexible Multi-task Networks by Learning Parameter Allocation
• Improving Explainable Recommendations with Synthetic Reviews
• Evaluating neural network explanation methods using hybrid documents and morphological agreement
• MARS: Memory Attention-Aware Recommender System
• Deep Automated Multi-task Learning
• Small Towers Make Big Differences
• Boosting Share Routing for Multi-task Learning
• New Item Consumption Prediction Using Deep Learning
• On Sampling Strategies for Neural Network-based Collaborative Filtering
• Effective Visualization and Analysis of Recommender Systems
• Implicit Feedback Deep Collaborative Filtering Product Recommendation System
• LCMR: Local and Centralized Memories for Collaborative Filtering with Unstructured Text
• Developing Multi-Task Recommendations with Long-Term Rewards via Policy Distilled Reinforcement Learning
• ESANS: Effective and Semantic-Aware Negative Sampling for Large-Scale Retrieval Systems
• SeER: An Explainable Deep Learning MIDI-based Hybrid Song Recommender System
• Interest-Related Item Similarity Model Based on Multimodal Data for Top-N Recommendation
• Personalized Neural Embeddings for Collaborative Filtering with Text
• Graphing else matters: exploiting aspect opinions and ratings in explainable graph-based recommendations
• Photos Are All You Need for Reciprocal Recommendation in Online Dating
• Multi-Faceted Hierarchical Multi-Task Learning for a Large Number of Tasks with Multi-dimensional Relations
• Multi-Channel Sequential Behavior Networks for User Modeling in Online Advertising
• Towards Explainable Fact Checking
• Hybrid Interest Modeling for Long-tailed Users
• Deep Latent Emotion Network for Multi-Task Learning
• Distilling neural networks into skipgram-level decision lists
• Pedestrian Tracking with Gated Recurrent Units and Attention Mechanisms