【论文笔记】Self-regulation: Employing a Generative Adversarial Network to Improve Event Detection

Introduction

Challenge 1: common words, ambiguous words and pronouns frequently used in the event that they are much harder to detect

Generality – taken home <Transport>
Ambiguty 1 – campaign in Iraq <Attack>
Ambiguty 2 – political campaign <Elect>
Coreference – Either its bad or good <Marry>

Challenge 2: based on neural network approach affected more features from the false, where false information to potential features are designated with similar events in the semantics, but in fact is not the case

Prison authorities have given the nod for Anwar to be taken home later in the afternoon.
Trigger: taken. Event Type: Transport

Experimental setup and results

Datasets: ACE2005, TAC-KBP2015

Experimental results

1. Trigger recognition
   
2. Event Category
   
3. embedding type
   
4. Suitability