MMALFM: Explainable Recommendation by Leveraging Reviews and Images

Abstract

Although the latent factor model achieves good accuracy in rating prediction, it suffers from many problems including cold-start, non-transparency, and suboptimal results for individual user-item pairs. In this paper, we exploit textual reviews and item images together with ratings to tackle these limitations. Specifically, we first apply a proposed multi-modal aspect-aware topic model (MATM) on text reviews and item images to model users' preferences and items' features from different aspects, and also estimate the aspect importance of a user towards an item. Then the aspect importance is integrated into a novel aspect-aware latent factor model (ALFM), which learns user's and item's latent factors based on ratings. In particular, ALFM introduces a weight matrix to associate those latent factors with the same set of aspects in MATM, such that the latent factors could be used to estimate aspect ratings. Finally, the overall rating is computed via a linear combination of the aspect ratings, which are weighted by the corresponding aspect importance. To this end, our model could alleviate the data sparsity problem and gain good interpretability for recommendation. Besides, every aspect rating is weighted by its aspect importance, which is dependent on the targeted user's preferences and the targeted item's features. Therefore, it is expected that the proposed method can model a user's preferences on an item more accurately for each user-item pair. Comprehensive experimental studies have been conducted on the Yelp 2017 Challenge dataset and Amazon product datasets to demonstrate the effectiveness of our method.

arXiv commentThis paper has been accepted by Transactions on Information Systems. arXiv admin note: substantial text overlap with arXiv:1802.07938

References in corpus (8)

• BPR: Bayesian Personalized Ranking from Implicit Feedback
• Ups and Downs: Modeling the Visual Evolution of Fashion Trends with One-Class Collaborative Filtering
• TransNets: Learning to Transform for Recommendation
• Aspect-Aware Latent Factor Model: Rating Prediction with Ratings and Reviews
• Neural Collaborative Filtering
• Vista: A Visually, Socially, and Temporally-aware Model for Artistic Recommendation
• Sherlock: Sparse Hierarchical Embeddings for Visually-aware One-class Collaborative Filtering
• Large Scale Visual Recommendations From Street Fashion Images

Cited by in corpus (28)

• Denoising Implicit Feedback for Recommendation
• Estimation-Action-Reflection: Towards Deep Interaction Between Conversational and Recommender Systems
• Attentive Aspect Modeling for Review-aware Recommendation
• User Diverse Preference Modeling by Multimodal Attentive Metric Learning
• Learning to Recommend with Multiple Cascading Behaviors
• Click-Through Rate Prediction with Multi-Modal Hypergraphs
• Formalizing Multimedia Recommendation through Multimodal Deep Learning
• Leave No User Behind: Towards Improving the Utility of Recommender Systems for Non-mainstream Users
• DBRec: Dual-Bridging Recommendation via Discovering Latent Groups
• Don't Explain without Verifying Veracity: An Evaluation of Explainable AI with Video Activity Recognition
• Hierarchical Hyperedge Embedding-based Representation Learning for Group Recommendation
• Adversarial Training Towards Robust Multimedia Recommender System
• Price-aware Recommendation with Graph Convolutional Networks
• Multi-criteria Rating and Review based Recommendation Model
• Dual Preference Distribution Learning for Item Recommendation
• Modeling Embedding Dimension Correlations via Convolutional Neural Collaborative Filtering
• Dissecting users' needs for search result explanations
• Quantifying and Alleviating the Language Prior Problem in Visual Question Answering
• Interest-aware Message-Passing GCN for Recommendation
• Large-scale Benchmarks for Multimodal Recommendation with Ducho
• Category Aware Explainable Conversational Recommendation
• Multi-Feature Discrete Collaborative Filtering for Fast Cold-start Recommendation
• Personalized Hashtag Recommendation for Micro-videos
• Boosting the Rating Prediction with Click Data and Textual Contents
• Reliable Shot Identification for Complex Event Detection via Visual-Semantic Embedding
• Hierarchical Text Interaction for Rating Prediction
• Attentive Long Short-Term Preference Modeling for Personalized Product Search
• Visually Aware Skip-Gram for Image Based Recommendations