DERA: Dense Entity Retrieval for Entity Alignment in Knowledge Graphs

TL;DR

This paper introduces DERA, a dense entity retrieval framework utilizing language models for improved entity alignment across knowledge graphs, achieving state-of-the-art results in both cross-lingual and monolingual settings.

Contribution

DERA is the first to use a dense retrieval approach with language models for entity alignment, enabling uniform encoding of entity features and better candidate ranking.

Findings

DERA outperforms existing methods on multiple datasets.

The approach effectively integrates structural and attribute information.

State-of-the-art accuracy achieved in cross-lingual and monolingual EA.

Abstract

Entity Alignment (EA) aims to match equivalent entities in different Knowledge Graphs (KGs), which is essential for knowledge fusion and integration. Recently, embedding-based EA has attracted significant attention and many approaches have been proposed. Early approaches primarily focus on learning entity embeddings from the structural features of KGs, defined by relation triples. Later methods incorporated entities' names and attributes as auxiliary information to enhance embeddings for EA. However, these approaches often used different techniques to encode structural and attribute information, limiting their interaction and mutual enhancement. In this work, we propose a dense entity retrieval framework for EA, leveraging language models to uniformly encode various features of entities and facilitate nearest entity search across KGs. Alignment candidates are first generated through entity retrieval, which are subsequently reranked to determine the final alignments. We conduct comprehensive experiments on both cross-lingual and monolingual EA datasets, demonstrating that our approach achieves state-of-the-art performance compared to existing EA methods.