This page provides the code and datasets of the ACL 2017 paper: Towards a Seamless Integration of Word Senses into Downstream NLP Applications
The followings are commands to get started:
python topic_categorization/__main__.py bbc word 5
python topic_categorization/__main__.py bbc supersense_wn 5
python topic_categorization/__main__.py bbc wn 5
python topic_categorization/__main__.py bbc supersense_bn 5
python topic_categorization/__main__.py bbc bn 5
python topic_categorization/__main__.py ohsumed word 23 --vocabsize 40000
python topic_categorization/__main__.py ohsumed supersense_wn 23 --vocabsize 40000
python topic_categorization/__main__.py ohsumed wn 23 --vocabsize 40000
python topic_categorization/__main__.py ohsumed supersense_bn 23 --vocabsize 40000
python topic_categorization/__main__.py ohsumed bn 23 --vocabsize 40000
python sentiment_analysis/__main__.exp.py PL04 word
python sentiment_analysis/__main__.exp.py PL04 wn
python sentiment_analysis/__main__.exp.py PL04 supersense_wn
python sentiment_analysis/__main__.exp.py PL04 bn
python sentiment_analysis/__main__.exp.py PL04 supersense_bn
Please note that these embeddings live in the same semantic space of Word2vec trained on the Google News dataset.
- Please read the paper for more information.
@InProceedings{pilehvar-EtAl:2017:Long,
author = {Pilehvar, Mohammad Taher and Camacho-Collados, Jose and Navigli, Roberto and Collier, Nigel},
title = {Towards a Seamless Integration of Word Senses into Downstream NLP Applications},
booktitle = {Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
month = {July},
year = {2017},
address = {Vancouver, Canada},
publisher = {Association for Computational Linguistics},
pages = {1857--1869},
url = {http://aclweb.org/anthology/P17-1170}
}
Have you had any questions, please contact us at