HyperStar: A System for Extraction of Hypernyms based on Word Embeddings

This is the implementation of the projection learning approach for learning word subsumptions, i.e., hyponyms and hypernyms, originally proposed by Fu et al. (2014). The approach requires pre-trained word embeddings in the word2vec format and the list of subsumption examples to learn the projection matrix. This implementation uses TensorFlow. The system is available at Github.

Citation

In case this software, the study or the dataset was useful for you, please cite the following paper.

Ustalov, D., Arefyev, N., Biemann, C., Panchenko, A.: Negative Sampling Improves Hypernymy Extraction Based on Projection Learning. In: Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 2, Short Papers, Valencia, Spain, Association for Computational Linguistics (April 2017) 543–550

@inproceedings{Ustalov:17:eacl,
author = {Ustalov, Dmitry and Arefyev, Nikolay and Biemann, Chris and Panchenko, Alexander},
title = {{Negative Sampling Improves Hypernymy Extraction Based on Projection Learning}},
booktitle = {Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 2, Short Papers},
month = {April},
year = {2017},
address = {Valencia, Spain},
publisher = {Association for Computational Linguistics},
pages = {543--550},
isbn = {978-1-945626-35-7},
url = {http://www.aclweb.org/anthology/E17-2087},
language = {english},
}