Abstract
Previous incremental parsers have used monotonic state transitions. However, transitions can be made to revise previous decisions quite naturally, based on further information. We show that a simple adjustment to the Arc-Eager transition system to relax its monotonicity constraints can improve accuracy, so long as the training data includes examples of mistakes for the non-monotonic transitions to repair. We evaluate the change in the context of a state-of-the-art system, and obtain a statistically significant improvement (p < 0.001) on the English evaluation and 5/10 of the CoNLL languages.
| Original language | English |
|---|---|
| Title of host publication | CoNLL 2013 - 17th Conference on Computational Natural Language Learning, Proceedings |
| Publisher | Association for Computational Linguistics (ACL) |
| Pages | 163-172 |
| Number of pages | 10 |
| ISBN (Electronic) | 9781937284701 |
| State | Published - 2013 |
| Event | 17th Conference on Computational Natural Language Learning, CoNLL 2013 - Sofia, Bulgaria Duration: 8 Aug 2013 → 9 Aug 2013 |
Publication series
| Name | CoNLL 2013 - 17th Conference on Computational Natural Language Learning, Proceedings |
|---|
Conference
| Conference | 17th Conference on Computational Natural Language Learning, CoNLL 2013 |
|---|---|
| Country/Territory | Bulgaria |
| City | Sofia |
| Period | 8/08/13 → 9/08/13 |
Bibliographical note
Publisher Copyright:© 2013 Association for Computational Linguistics.
Funding
The authors would like to thank the anonymous reviewers for their valuable comments. This research was supported under the Australian Research Council’s Discovery Projects funding scheme (project numbers DP110102506 and DP110102593).
| Funders | Funder number |
|---|---|
| Australian Research Council | DP110102506, DP110102593 |