Abstract

Syntactic relations are broadly used in many NLP tasks. For event detection, syntactic relation representations based on dependency tree can better capture the interrelations between candidate trigger words and related entities than sentence representations. But, existing studies only use first-order syntactic relations (i.e., the arcs) in dependency trees to identify trigger words. For this reason, this paper proposes a new method for event detection, which uses a dependency tree based graph convolution network with aggregative attention to explicitly model and aggregate multi-order syntactic representations in sentences. Experimental comparison with state-of-the-art baselines shows the superiority of the proposed method.

Highlights

  • As an important information extraction task, event detection aims to find event mentions with specific event types from given texts

  • Among existing methods for event detection, sequence based ones (e.g., Chen et al (2015); Nguyen et al (2016)) only use the given sentences, which suffer from the low efficiency problem in capturing long-range dependency; On the contrary, dependency tree based methods utilize the syntactic relations in the dependency tree of a given sentence to more effectively capture the interrelation between each candidate trigger word and its related entities or other triggers

  • Since each word can only be updated by its neighbors in the dependency tree through dependency arcs, following the previous methods (Liu et al, 2018), we employ a Bidirectional Long-Short Term Memory network (BiLSTM) (Hochreiter and Schmidhuber, 1997) to encode X with its context as P = p1, p2, ..., pn, which will be used as the input of the multi-order Graph Attention Network (GAT)

Read more

Summary

Introduction

As an important information extraction task, event detection aims to find event mentions with specific event types from given texts. As the number of GCN layers increases, the representations of neighboring words in the dependency tree will get more and more similar, since they all are calculated via those of their neighbors in the dependency tree This is the so-called over-smoothing problem (Zhou et al, 2018), which damages the diversity of the representations of neighboring words. To overcome this problem, this paper proposes a Multi-Order Graph Attention Network based method for Event Detection, called MOGANED. This paper proposes a Multi-Order Graph Attention Network based method for Event Detection, called MOGANED It utilizes both first-order syntactic graph and highorder syntactic graphs to explicitly model multiorder representations of candidate trigger words. It should be mentioned that, to the best of our knowledge, this is the first study applying GAT to event detection

The Proposed Model
Word Encoding
Multi-Order Graph Attention Network
Attention Aggregation
Bias Loss Function
Experiment Settings
Method
Overall Performance
Ablation Study
Findings
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call