Additive Compositionality of Word Vectors

Yeon Seonwoo, Sungjoon Park, Dongkwan Kim, Alice Oh


Abstract
Additive compositionality of word embedding models has been studied from empirical and theoretical perspectives. Existing research on justifying additive compositionality of existing word embedding models requires a rather strong assumption of uniform word distribution. In this paper, we relax that assumption and propose more realistic conditions for proving additive compositionality, and we develop a novel word and sub-word embedding model that satisfies additive compositionality under those conditions. We then empirically show our model’s improved semantic representation performance on word similarity and noisy sentence similarity.
Anthology ID:
D19-5551
Volume:
Proceedings of the 5th Workshop on Noisy User-generated Text (W-NUT 2019)
Month:
November
Year:
2019
Address:
Hong Kong, China
Editors:
Wei Xu, Alan Ritter, Tim Baldwin, Afshin Rahimi
Venue:
WNUT
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
387–396
Language:
URL:
https://aclanthology.org/D19-5551
DOI:
10.18653/v1/D19-5551
Bibkey:
Cite (ACL):
Yeon Seonwoo, Sungjoon Park, Dongkwan Kim, and Alice Oh. 2019. Additive Compositionality of Word Vectors. In Proceedings of the 5th Workshop on Noisy User-generated Text (W-NUT 2019), pages 387–396, Hong Kong, China. Association for Computational Linguistics.
Cite (Informal):
Additive Compositionality of Word Vectors (Seonwoo et al., WNUT 2019)
Copy Citation:
PDF:
https://aclanthology.org/D19-5551.pdf