Contextual Semantic-Guided Entity-Centric GCN for Relation Extraction

Relation extraction tasks aim to predict potential relations between entities in a target sentence. As entity mentions have ambiguity sentences, some important contextual information can guide the semantic representation of improve accuracy relation extraction. However, most existing models ignore guidance and treat textual context sentence equally. This results low-accuracy extractions. To address this problem, we propose semantic-guided entity-centric graph convolutional network (CEGCN) model that enables obtain for more accurate relational representations. develops self-attention enhanced neural concentrate on importance relevance different words information. Then, employ dependency tree with as global nodes add virtual edges construct an logical adjacency matrix (ELAM). enable aggregate one-layer GCN calculation. The experimental TACRED SemEval-2010 Task 8 datasets show our efficiently use enrich representations outperform previous models.