Variants of Vector Space Reductions for Predicting the Compositionality of English Noun Compounds
Abstract
Predicting the degree of compositionality of noun compounds is a crucial ingredient for lexicography and NLP applications, to know whether the compound should be treated as a whole, or through its constituents. Computational approaches for an automatic prediction typically represent compounds and their constituents within a vector space to have a numeric relatedness measure for the words. This paper provides a systematic evaluation of using different vector-space reduction variants for the prediction. We demonstrate that Word2vec and nouns-only dimensionality reductions are the most successful and stable vector space reduction variants for our task.- Anthology ID:
- 2020.winlp-1.13
- Volume:
- Proceedings of the The Fourth Widening Natural Language Processing Workshop
- Month:
- July
- Year:
- 2020
- Address:
- Seattle, USA
- Venue:
- WiNLP
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 51–54
- Language:
- URL:
- https://aclanthology.org/2020.winlp-1.13
- DOI:
- 10.18653/v1/2020.winlp-1.13
- Cite (ACL):
- Pegah Alipoormolabashi and Sabine Schulte im Walde. 2020. Variants of Vector Space Reductions for Predicting the Compositionality of English Noun Compounds. In Proceedings of the The Fourth Widening Natural Language Processing Workshop, pages 51–54, Seattle, USA. Association for Computational Linguistics.
- Cite (Informal):
- Variants of Vector Space Reductions for Predicting the Compositionality of English Noun Compounds (Alipoormolabashi & Schulte im Walde, WiNLP 2020)