Transition-Based Syntactic Linearization with Lookahead Features.
- ,
- Yue Zhang,
- Manish Shrivastava
- Kohli Center on Intelligent Systems,
- Singapore University of Technology and Design
Research Output:
Conference Article in Proceeding or Book/Report chapter
Article in proceedings
Peer-review
Open access
Publication Information
Output type
Research Output:
Conference Article in Proceeding or Book/Report chapter
Article in proceedings
Peer-reviewOriginal language
EnglishPages from-to (Number of pages)
Pages 488-493Publication milestones
- Published - 2016
Publication status
Published - 2016
Publisher
Association for Computational Linguistics, United StatesPublication IDs
- Scopus: 84994137698
Host publication title
Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language TechnologiesAbstract
It has been shown that transition-based methods can be used for syntactic word ordering and tree linearization, achieving significantly faster speed compared with traditional best-first methods. State-of-the-art transitionbased models give competitive results on abstract word ordering and unlabeled tree linearization, but significantly worse results onlabeled tree linearization. We demonstrate that
the main cause for the performance bottleneck is the sparsity of SHIFT transition actions rather than heavy pruning. To address this issue, we propose a modification to the standard transition-based feature structure, which
reduces feature sparsity and allows lookahead features at a small cost to decoding efficiency. Our model gives the best reported accuracies
on all benchmarks, yet still being over 30 times faster compared with best-first-search
the main cause for the performance bottleneck is the sparsity of SHIFT transition actions rather than heavy pruning. To address this issue, we propose a modification to the standard transition-based feature structure, which
reduces feature sparsity and allows lookahead features at a small cost to decoding efficiency. Our model gives the best reported accuracies
on all benchmarks, yet still being over 30 times faster compared with best-first-search
Publication metrics
PlumX, opens in new tab
Citations
13
Captures
81