No description, website, or topics provided.
Branch: master
Clone or download
Latest commit bf57b20 Jan 28, 2019
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
data Update README.md May 20, 2018
README.md Update README.md Jan 28, 2019
batcher.py Update batcher.py Jul 15, 2018
experiments.conf Update experiments.conf May 20, 2018
model.py Update model.py May 20, 2018
preprocess.py Update preprocess.py Jun 10, 2018

README.md

Robust Lexical Features for Improved Neural Network Named-Entity Recognition

This repository contains the source code for the NER system presented in the following research publication (link)

Abbas Ghaddar and Philippe Langlais 
Robust Lexical Features for Improved Neural Network Named-Entity Recognition
In Proceedings of the 27th International Conference on Computational Linguistics (COLING 2018)

Requirements

  • python 3.6
  • tensorflow>=1.6
  • pyhocon (for parsing the configurations)

Prepare the Data

  1. Download the data from here and unzip the files in data directory.

  2. Change the raw_path variables for conll and ontonotes datasets in experiments.config file to path/to/conll-2003 and path/to/conll-2012/v4/data respectively. For conll dataset please rename eng.train eng.testa eng.testb files to conll.train.txt conll.dev.txt conll.test.txt respectively.

  3. Run:

$ python preprocess.py dataset_name[conll|ontonotes]

Training

Once the data preprocessing is completed, you can train and test a model with:

$ python model.py dataset_name[conll|ontonotes]

Generate LS embeddings

The following link contains the model, entity type vocab and code to generate LS embeddings for any word.

Citation

Please cite the following paper when using our code:

@InProceedings{ghaddar2018coling,
  title={Robust Lexical Features for Improved Neural Network Named-Entity Recognition},
  author={Ghaddar, Abbas	and Langlais, Phillippe},
  booktitle = {{COLING} 2018, 27th International Conference on Computational Linguistics},
  pages     = {1896--1907},
  year      = {2018}
}