Abstract
Three broad approaches have been attempted to combine distributional and structural/symbolic aspects to construct meaning representations: a) injecting linguistic features into distributional representations, b) injecting distributional features into symbolic representations or c) combining structural and distributional features in the final representation. This work focuses on an example of the third and less studied approach: it extends the Graphical Knowledge Representation (GKR) to include distributional features and proposes a division of semantic labour between the distributional and structural/symbolic features. We propose two extensions of GKR that clearly show this division and empirically test one of the proposals on an NLI dataset with hard compositional pairs.- Anthology ID:
- W19-3305
- Volume:
- Proceedings of the First International Workshop on Designing Meaning Representations
- Month:
- August
- Year:
- 2019
- Address:
- Florence, Italy
- Editors:
- Nianwen Xue, William Croft, Jan Hajic, Chu-Ren Huang, Stephan Oepen, Martha Palmer, James Pustejovksy
- Venue:
- DMR
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 44–55
- Language:
- URL:
- https://aclanthology.org/W19-3305
- DOI:
- 10.18653/v1/W19-3305
- Cite (ACL):
- Aikaterini-Lida Kalouli, Richard Crouch, and Valeria de Paiva. 2019. GKR: Bridging the Gap between Symbolic/structural and Distributional Meaning Representations. In Proceedings of the First International Workshop on Designing Meaning Representations, pages 44–55, Florence, Italy. Association for Computational Linguistics.
- Cite (Informal):
- GKR: Bridging the Gap between Symbolic/structural and Distributional Meaning Representations (Kalouli et al., DMR 2019)
- PDF:
- https://preview.aclanthology.org/nschneid-patch-3/W19-3305.pdf
- Code
- kkalouli/GKR_semantic_parser
- Data
- SNLI