Hierarchical Embeddings for Hypernymy Detection and Directionality
Abstract
We present a novel neural model HyperVec to learn hierarchical embeddings for hypernymy detection and directionality. While previous embeddings have shown limitations on prototypical hypernyms, HyperVec represents an unsupervised measure where embeddings are learned in a specific order and capture the hypernymhyponym distributional hierarchy. Moreover, our model is able to generalize over unseen hypernymy pairs, when using only small sets of training data, and by mapping to other languages. Results on benchmark datasets show that HyperVec outperforms both stateoftheart unsupervised measures and embedding models on hypernymy detection and directionality, and on predicting graded lexical entailment.
Cite
@article{arxiv.1707.07273,
title = {Hierarchical Embeddings for Hypernymy Detection and Directionality},
author = {Kim Anh Nguyen and Maximilian Köper and Sabine Schulte im Walde and Ngoc Thang Vu},
journal= {arXiv preprint arXiv:1707.07273},
year = {2017}
}
Comments
11 pages, accepted as long paper at EMNLP 2017