Document-level Core Events Extraction based on QA

Abstract

Document-level core events extraction model based on QA is provided in this paper. Question with prior information is constructed input into the BTBiLSTM model. Firstly, extract events from documents. Secondly finds event arguments from the text to be extracted. Lastly fuse events to select core events. The experimental results on ACE2005 Chinese corpus show that the event argument role question with prior information has a good effect on event argument extraction. Compared with existing method, our method in this paper realizes cross sentences events arguments extraction, it also applicable to documents without headline.