Explainable Recommendation with Personalized Review Retrieval and Aspect Learning

TL;DR

This paper introduces ERRA, a novel explainable recommendation model that enhances explanation accuracy through review retrieval and aspect learning, significantly improving prediction and explanation quality on multiple datasets.

Contribution

The paper proposes ERRA, combining review retrieval and aspect learning to generate more accurate explanations and better capture user preferences in recommendation systems.

Findings

ERRA outperforms state-of-the-art baselines in prediction accuracy.

ERRA achieves a 15.8% improvement in explanation quality on TripAdvisor.

The model effectively captures user preferences through aspect selection.

Abstract

Explainable recommendation is a technique that combines prediction and generation tasks to produce more persuasive results. Among these tasks, textual generation demands large amounts of data to achieve satisfactory accuracy. However, historical user reviews of items are often insufficient, making it challenging to ensure the precision of generated explanation text. To address this issue, we propose a novel model, ERRA (Explainable Recommendation by personalized Review retrieval and Aspect learning). With retrieval enhancement, ERRA can obtain additional information from the training sets. With this additional information, we can generate more accurate and informative explanations. Furthermore, to better capture users' preferences, we incorporate an aspect enhancement component into our model. By selecting the top-n aspects that users are most concerned about for different items, we can model user representation with more relevant details, making the explanation more persuasive. To verify the effectiveness of our model, extensive experiments on three datasets show that our model outperforms state-of-the-art baselines (for example, 3.4% improvement in prediction and 15.8% improvement in explanation for TripAdvisor).