Event detection with trigger-aware lattice neural network

Abstract

Event detection (ED) aims to locate trigger words in raw text and then classify them into correct event types. In this task, neural network based models became mainstream in recent years. However, two problems arise when it comes to languages without natural delimiters, such as Chinese. First, word-based models severely suffer from the problem of word-trigger mismatch, limiting the performance of the methods. In addition, even if trigger words could be accurately located, the ambiguity of polysemy of triggers could still affect the trigger classification stage. To address the two issues simultaneously, we propose the Trigger-aware Lattice Neural Network (TLNN). (1) The framework dynamically incorporates word and character information so that the trigger-word mismatch issue can be avoided. (2) Moreover, for polysemous characters and words, we model all senses of them with the help of an external linguistic knowledge base, so as to alleviate the problem of ambiguous triggers. Experiments on two benchmark datasets show that our model could effectively tackle the two issues and outperforms previous state-of-the-art methods significantly, giving the best results. The source code of this paper can be obtained from https://github.com/thunlp/TLNN.