Contrastive learning boosts explainable product recommendations
Contrastive Learning for Aspect Representation towards Explainable Recommendation
Information RetrievalArtificial Intelligence
Summary
Recommender systems suggest products or services that users might like, but understanding why can be hard. This work introduces a method that learns from both user ratings and the detailed opinions in reviews to improve recommendations. It uses a technique called contrastive learning to better capture what users prefer about specific product aspects. This combined approach not only improves recommendation accuracy but also helps generate clear explanations for the suggestions. Tests on multiple benchmark datasets show this method works better than previous ones.
What this means in practice
- •For ecommerce developers: Enhance online shopping platforms by integrating improved recommendation accuracy together with user-friendly explanations based on review aspects and ratings.$Commercial implications: This method enables ecommerce systems to offer buyers better suggestions with clear reasons, which can increase sales and user engagement.
- •For digital marketing teams: Create targeted marketing strategies by understanding which product aspects drive user preferences through combined rating and review analysis.
Authors
Emrul Hasan, Chen Ding
Abstract
In this work, we propose a novel recommendation model, CLARER (Contrastive Learning for Aspect Representation towards Explainable Recommendation) that integrates aspect features learned from textual reviews with rating information to improve the accuracy and explainability of recommendations. Our proposed framework learns user and item representations by combining rating-based features and aspect-based features from reviews. Specifically, rating-based features are learned through a multi-layer perceptron (MLP) model, while aspect-specific review representations are learned using a transformer encoder to capture the semantic information and contrastive learning to better distinguish user preferences. To provide explanations, we train a transformer decoder, using the final representations of users and items from both rating and aspect-based features as context. Experimental results in three benchmark data sets demonstrate that our model achieves superior performance compared to baseline methods in both recommendation (accuracy) and explanation generation.