Improving Semantic Composition with Offset Inference

Thomas Kober, Julie Weeds, Jeremy Reffin, David Weir


Abstract
Count-based distributional semantic models suffer from sparsity due to unobserved but plausible co-occurrences in any text collection. This problem is amplified for models like Anchored Packed Trees (APTs), that take the grammatical type of a co-occurrence into account. We therefore introduce a novel form of distributional inference that exploits the rich type structure in APTs and infers missing data by the same mechanism that is used for semantic composition.
Anthology ID:
P17-2069
Volume:
Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)
Month:
July
Year:
2017
Address:
Vancouver, Canada
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
433–440
Language:
URL:
https://aclanthology.org/P17-2069
DOI:
10.18653/v1/P17-2069
Bibkey:
Cite (ACL):
Thomas Kober, Julie Weeds, Jeremy Reffin, and David Weir. 2017. Improving Semantic Composition with Offset Inference. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), pages 433–440, Vancouver, Canada. Association for Computational Linguistics.
Cite (Informal):
Improving Semantic Composition with Offset Inference (Kober et al., ACL 2017)
Copy Citation:
PDF:
https://preview.aclanthology.org/auto-file-uploads/P17-2069.pdf
Poster:
 P17-2069.Poster.pdf
Software:
 P17-2069.Software.tgz
Code
 tttthomasssss/acl2017