BNET-DS

Abstract: Distant Supervision (DS) is a popular tech- nique for developing relation extractors start- ing with limited supervision. Our contri- butions in this paper are threefold. Firstly, we propose three novel models for distantly- supervised relation extraction: (1) a Bi-GRU based word attention model (BGWA), (2) an entity-centric attention model (EA), and (3) and a combination model (BNET-DS) which jointly trains and combines multiple comple- mentary models for improved relation extrac- tion. Secondly, we introduce GDS, a new distant supervision dataset for relation extrac- tion. GDS removes test data noise present in all previous distance supervision benchmark datasets, making credible automatic evalu- ation possible. Thirdly, through extensive experiments on multiple real-world datasets, we demonstrate effectiveness of the proposed methods.