Skip to content

WangHanLinHenry/event_detection_without_triggers

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Event Detection without Triggers (NAACL2019)

This repository provides the code for the work in NAACL2019: Event Detection without Triggers.

Because of the copyright issue of ACE2005 Corpus, we can not release the corpus. For test, we give 10 samples in data/test_corpus_10.txt.

Each line represents a testing sample, whose format is as follows:

w1 e1 \t w2 e2 \t ... ... wn en \t evt1 evt2 ... evtm

where, [w1, w2, ..., wn] are tokens of a testing sentence, [e1, e2, ..., en] are the corresponding entity type of each token, [evt1, evt2, ..., evtm] are the types of events mentioned in this sentence (if m is 0, this block will be replaced with a single 'NEGATIVE' label).

We provide a trained model, which can be downloaded here: model files.

You can run this code to evaluate the trained model using the following command:

python run_model.py evaluation 

or train the model using your own traininng corpus:

python run_model.py train 

Required running environment:
1. python 2.7
2. tensorflow 1.4 or higher

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%