Abstract:Explanations accompanied by a recommendation can assist users in understanding the decision made by recommendation systems, which in turn increases a user's confidence and trust in the system. Recently, research has focused on generating natural language explanations in a human-readable format. Thus far, the proposed approaches leverage item reviews written by users, which are often subjective, sparse in language, and unable to account for new items that have not been purchased or reviewed before. Instead, we aim to generate fact-grounded recommendation explanations that are objectively described with item features while implicitly considering a user's preferences, based on the user's purchase history. To achieve this, we propose a knowledge graph (KG) approach to natural language explainable recommendation. Our approach draws on user-item features through a novel collaborative filtering-based KG representation to produce fact-grounded, personalized explanations, while jointly learning user-item representations for recommendation scoring. Experimental results show that our approach consistently outperforms previous state-of-the-art models on natural language explainable recommendation.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the limitations of natural - language explanations in existing recommendation systems. Specifically, existing methods mainly rely on user - provided reviews to generate explanations. These reviews are often subjective, sparsely expressed in language, and unable to provide reasonable explanations for new products. These problems make it difficult for users to understand the reasons for the recommendations, thus affecting users' trust and user experience. Therefore, this paper proposes a method based on the Knowledge Graph (KG) aimed at generating objective, specific, and data - driven recommendation explanations while considering users' purchase history to improve the transparency and credibility of the recommendation system.
### Main contributions of the paper:
1. **Generate objective explanations**: By using the knowledge graph of products, the generated explanations are more objective and specific and can better describe the characteristics of products.
2. **Handle new products**: Unlike methods that rely on user reviews, this method can generate explanations for new products that have not been purchased or reviewed, solving the cold - start problem.
3. **Personalized recommendation**: Combine users' purchase history to generate personalized recommendation explanations, improving users' satisfaction and trust.
4. **Multi - task learning**: Simultaneously perform recommendation scoring and explanation generation, and improve the overall performance of the model through joint training.
### Method overview:
- **Input representation**: Users are represented by their purchase history, and products are represented by their knowledge graph.
- **Encoder**: Adopt the Transformer architecture, including the global attention mechanism and the user - product graph attention mechanism to capture the topological information of the user - product graph.
- **Output module**: Generate recommendation scores and natural - language explanations, and optimize the model through multi - task learning.
### Experimental results:
- **Explanation quality**: On multiple evaluation metrics (such as BLEU, ROUGE, USR, EC), the proposed model significantly outperforms the baseline model, especially outstanding in terms of generation diversity and entity coverage.
- **Recommendation performance**: On the recommendation task, the RMSE and MAE indicators of the model also perform well, indicating that this method has an advantage not only in explanation generation but also in recommendation accuracy.
### Conclusion:
The method proposed in this paper effectively solves the problems of subjectivity and sparsity of explanations in existing recommendation systems. By introducing the knowledge graph, more objective, specific, and personalized recommendation explanations are generated, improving users' trust and user experience.