Long-Distance Dependencies Don’t Have to Be Long: Simplifying through Provably (Approximately) Optimal Permutations

Rishi Bommasani


Abstract
Neural models at the sentence level often operate on the constituent words/tokens in a way that encodes the inductive bias of processing the input in a similar fashion to how humans do. However, there is no guarantee that the standard ordering of words is computationally efficient or optimal. To help mitigate this, we consider a dependency parse as a proxy for the inter-word dependencies in a sentence and simplify the sentence with respect to combinatorial objectives imposed on the sentence-parse pair. The associated optimization results in permuted sentences that are provably (approximately) optimal with respect to minimizing dependency parse lengths and that are demonstrably simpler. We evaluate our general-purpose permutations within a fine-tuning schema for the downstream task of subjectivity analysis. Our fine-tuned baselines reflect a new state of the art for the SUBJ dataset and the permutations we introduce lead to further improvements with a 2.0% increase in classification accuracy (absolute) and a 45% reduction in classification error (relative) over the previous state of the art.
Anthology ID:
P19-2012
Volume:
Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics: Student Research Workshop
Month:
July
Year:
2019
Address:
Florence, Italy
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
89–99
Language:
URL:
https://aclanthology.org/P19-2012
DOI:
10.18653/v1/P19-2012
Bibkey:
Cite (ACL):
Rishi Bommasani. 2019. Long-Distance Dependencies Don’t Have to Be Long: Simplifying through Provably (Approximately) Optimal Permutations. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics: Student Research Workshop, pages 89–99, Florence, Italy. Association for Computational Linguistics.
Cite (Informal):
Long-Distance Dependencies Don’t Have to Be Long: Simplifying through Provably (Approximately) Optimal Permutations (Bommasani, ACL 2019)
Copy Citation:
PDF:
https://aclanthology.org/P19-2012.pdf