Abstract

Entity alignment aims to match synonymous entities across different knowledge graphs, which is a fundamental task for knowledge integration. Recently, researchers have devoted to leveraging rich information within relations to enhance entity alignment. They explicitly incorporate relations in entity representation and alignment, demonstrating remarkable results. However, affected by the semantic assumptions from early works, these works represent a relation by combining all the entities it connects, ignoring the semantic independence between entity and relation. Moreover, since these works perform alignment by comparing embedding similarity, they fail to consider a graph level alignment and tend to find local false correspondences. In this paper, we propose Entity and Relation Matching Consensus (ERMC), a two-stage matching schema based on graph matching consensus that jointly models and aligns entities and relations and retains their semantic independence at the same time. In the first stage, we design a bidirectional relation-aware graph convolutional network to jointly learn entity and relation embeddings based on the triadic graph by a novel message passing mechanism. Then, we jointly align the entities and relations by computing a graph-level matching consensus. In the second stage, we introduce a refinement strategy to detect and correct false alignments in the first stage. Experimental results on three real-world multilingual datasets demonstrate that ERMC outperforms some state-of-the-art models on both entity alignment and relation alignment tasks.

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.