论文信息 - Characterizing the Errors of Data-Driven Dependency Parsing Models

Characterizing the Errors of Data-Driven Dependency Parsing Models

We present a comparative error analysis of the two dominant approaches in datadriven dependency parsing: global, exhaustive, graph-based models, and local, greedy, transition-based models. We show that, in spite of similar performance overall, the two models produce different types of errors, in a way that can be explained by theoretical properties of the two models. This analysis leads to new directions for parser development.

Joakim Nivre | Ryan T. McDonald | Joakim Nivre

[1] Hiroshi Maruyama,et al. Structural Disambiguation with Constraint Propagation , 1990, ACL.

[2] Joakim Nivre,et al. Pseudo-Projective Dependency Parsing , 2005, ACL.

[3] John Langford,et al. Search-based structured prediction , 2009, Machine Learning.

[4] Daniel Jurafsky,et al. Learning Syntactic Patterns for Automatic Hypernym Discovery , 2004, NIPS.

[5] Sabine Buchholz,et al. CoNLL-X Shared Task on Multilingual Dependency Parsing , 2006, CoNLL.

[6] Fernando Pereira,et al. Multilingual Dependency Analysis with a Two-Stage Discriminative Parser , 2006, CoNLL.

[7] Martha Palmer,et al. Synchronous Dependency Insertion Grammars: A Grammar Formalism for Syntax Based Statistical MT , 2004 .

[8] Joakim Nivre. Constraints on Non-Projective Dependency Parsing , 2006, EACL.

[9] Alon Lavie,et al. Parser Combination by Reparsing , 2006, NAACL.

[10] 菅山謙正,et al. Word Grammar 理論の研究 , 2005 .

[11] Joakim Nivre,et al. An Efficient Algorithm for Projective Dependency Parsing , 2003, IWPT.

[12] Fernando Pereira,et al. Non-Projective Dependency Parsing using Spanning Tree Algorithms , 2005, HLT.

[13] Koby Crammer,et al. Online Large-Margin Training of Dependency Parsers , 2005, ACL.

[14] Igor Mel’čuk,et al. Dependency Syntax: Theory and Practice , 1987 .

[15] Joakim Nivre,et al. Labeled Pseudo-Projective Dependency Parsing with Support Vector Machines , 2006, CoNLL.