Difference between revisions of "Koo and Collins ACL 2010"
From Cohen Courses
Jump to navigationJump to search| Line 11: | Line 11: | ||
This [[Category::paper]] presents a higher-order [[AddressesProblem::Dependency Parsing|dependency parser]] that can evaluate substructures containing three dependencies, using both sibling-style and grandchild-style interactions. The algorithms presented require only <math>O(n^4)</math> time and were evaluated on the [[UsesDataset::Penn Treebank]] and the [[UsesDataset::Prague Dependency Treebank]]. The implementation code was publicly released [http://groups.csail.mit.edu/nlp/dpo3/]. | This [[Category::paper]] presents a higher-order [[AddressesProblem::Dependency Parsing|dependency parser]] that can evaluate substructures containing three dependencies, using both sibling-style and grandchild-style interactions. The algorithms presented require only <math>O(n^4)</math> time and were evaluated on the [[UsesDataset::Penn Treebank]] and the [[UsesDataset::Prague Dependency Treebank]]. The implementation code was publicly released [http://groups.csail.mit.edu/nlp/dpo3/]. | ||
| − | Dependency parsing is defined as a search for the highest-scoring analysis of x: | + | Dependency parsing is defined as a search for the highest-scoring analysis of <math> x </math>: |
<math>y^*(x) = _{y\in Y(x)}^{argmax}\textrm{Score}(x,y)</math> | <math>y^*(x) = _{y\in Y(x)}^{argmax}\textrm{Score}(x,y)</math> | ||
Revision as of 22:18, 25 November 2011
Citation
Koo, T. and Collins, M. 2010. Efficient Third-Order Dependency Parsers. In Proceedings of ACL, pp. 1-11. Association for Computational Linguistics.
Online version
Summary
This paper presents a higher-order dependency parser that can evaluate substructures containing three dependencies, using both sibling-style and grandchild-style interactions. The algorithms presented require only time and were evaluated on the Penn Treebank and the Prague Dependency Treebank. The implementation code was publicly released [1].
Dependency parsing is defined as a search for the highest-scoring analysis of :
Experimental results
Bla bla.
Related papers
Bla bla.
