Deep Semantic Match Model for Entity Linking Using Knowledge Graph and Text

Abstract

In this paper, we address the problem of Entity Linking (EL) as aligning a textual mention to the referent entity in a knowledge base (e.g., Freebase). Most previous studies on EL mainly focus on designing various feature representations for the mentions and entities. However, these handcrafted features often ignore the internal meanings of words or entities, require tedious feature engineering and expensive computation and lack adaptability for different scenarios. In this paper, we propose a Deep Semantic Match Model (DSMM) for EL by using knowledge graph and descriptive text. Specifically, the DSMM applys bidirectional Long Short Term Memory Network (BiLSTM) with multi-granularities to match mentions with candidate entities from two aspects: surface form match by a character-level BiLSTM (char-LSTM) and semantic match based on the "structural" context of entities and the textual context of mentions by a word-level BiLSTM (word-LSTM). Experimental results on CoNLL benchmark dataset show that our proposed DSMM significantly outperforms existing baseline models for EL task.