Event Detection as Graph Parsing

Abstract

Event detection is a fundamental task in information extraction. Most previous approaches typically view event detection as a triggerbased classification problem, focusing on using syntactic dependency structure or external knowledge to boost the classification performance. To overcome the inherent issues with existing trigger classification based models, we propose a novel approach to event detection by formulating it as a graph parsing problem, which can explicitly model the multiple event correlations and naturally utilize the rich information conveyed by event type and subtype. Furthermore, to cope with data sparsity, we employ a pretrained sequence-tosequence (seq2seq) model to transduce an input sentence into an accurate event graph without the need for trigger words. Extensive experimental results on the public ACE2005 dataset show that, our approach outperforms all previous state-of-the-art models for event detection by a large margin, obtaining an improvement of 4.2% F1 score. The result is very encouraging since we achieve this with a conceptually simple seq2seq model; moreover, by extending the graph structure, this proposed architecture can be flexibly applied to more information extraction problems for sentences.

Key News Event Detection and Event Context Using Graphic Convolution, Clustering, and Summarizing Methods

“…For sentencelevel event detection tasks, using graph structures has become a trend. Xie et al [8] proposed a graph parsing method using a dependency-tree-based graph and convolutional neural networks. He et al [9] presented a self-constructed dependency and graph convolution network for event detection.…”

Seq2EG: a novel and effective event graph parsing approach for event extraction

“…This paper is a significant extension of our conference paper [16], which presents the first work to formulate event detection as a graph parsing task, and to introduce a novel generation-based method to predict event graph containing only event type nodes. In this paper, we further demonstrate the universality of the graph parsing framework by extending it to more complicated event extraction task, and propose some skillful strategies for the complete event graph linearization and an effective decoding algorithm to boost the generation performance.…”