Abstract
Many natural language processing tasks, e.g., coreference resolution and semantic role labeling, require selecting text spans and making decisions about them. A typical approach to such tasks is to score all possible spans and greedily select spans for task-specific downstream processing. This approach, however, does not incorporate any inductive bias about what sort of spans ought to be selected, e.g., that selected spans tend to be syntactic constituents. In this paper, we propose a novel grammar-based structured span selection model which learns to make use of the partial span-level annotation provided for such problems. Compared to previous approaches, our approach gets rid of the heuristic greedy span selection scheme, allowing us to model the downstream task on an optimal set of spans. We evaluate our model on two popular span prediction tasks: coreference resolution and semantic role labeling; and show improvements on both.- Anthology ID:
- 2022.naacl-main.189
- Volume:
- Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies
- Month:
- July
- Year:
- 2022
- Address:
- Seattle, United States
- Editors:
- Marine Carpuat, Marie-Catherine de Marneffe, Ivan Vladimir Meza Ruiz
- Venue:
- NAACL
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 2629–2641
- Language:
- URL:
- https://aclanthology.org/2022.naacl-main.189
- DOI:
- 10.18653/v1/2022.naacl-main.189
- Cite (ACL):
- Tianyu Liu, Yuchen Jiang, Ryan Cotterell, and Mrinmaya Sachan. 2022. A Structured Span Selector. In Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 2629–2641, Seattle, United States. Association for Computational Linguistics.
- Cite (Informal):
- A Structured Span Selector (Liu et al., NAACL 2022)
- PDF:
- https://preview.aclanthology.org/nschneid-patch-4/2022.naacl-main.189.pdf
- Code
- lyutyuh/structured-span-selector
- Data
- CoNLL-2012