Multi-Aspect Explainable Inductive Relation Prediction by Sentence Transformer
This work addresses the need for more reliable and interpretable relation predictions in knowledge graphs, offering incremental improvements over existing methods.
The paper tackles the problem of improving inductive and explainable relation prediction in knowledge graphs by filtering unreliable paths and proposing a sentence transformer model, achieving state-of-the-art performance in most test cases, including 4 out of 6 transductive and inductive cases and 11 out of 12 few-shot cases.
Recent studies on knowledge graphs (KGs) show that path-based methods empowered by pre-trained language models perform well in the provision of inductive and explainable relation predictions. In this paper, we introduce the concepts of relation path coverage and relation path confidence to filter out unreliable paths prior to model training to elevate the model performance. Moreover, we propose Knowledge Reasoning Sentence Transformer (KRST) to predict inductive relations in KGs. KRST is designed to encode the extracted reliable paths in KGs, allowing us to properly cluster paths and provide multi-aspect explanations. We conduct extensive experiments on three real-world datasets. The experimental results show that compared to SOTA models, KRST achieves the best performance in most transductive and inductive test cases (4 of 6), and in 11 of 12 few-shot test cases.