Recent Advances in Dependency Parsing

Qin Iris Wang

Yue Zhang

AT&T Interactive

Cambridge University

[email protected]

[email protected]

NAACL Tutorial, Los Angeles June 1, 2010

Part E: Other Recent Trends in Dependency Parsing Qin Iris Wang AT&T Interactive [email protected]

NAACL Tutorial, Los Angeles June 1, 2010

Outline   Part

A: introduction to dependency parsing   Part B: graph-based dependency parsing models   Part C: transition-based models   Part D: the combined models   Part E: other recent trends in dependency parsing      

Explore higher order features Use extra information source Better parsing strategies 2

NAACL 2010 Tutorial ------ Qin Iris Wang & Yue Zhang------June 1, 2010

Other Recent Trends in Dependency Parsing   Explore   Use      

higher order features

extra information sources

Raw data Bilingual data Linguistic rules

  Better

parsing strategies 3

NAACL 2010 Tutorial ------ Qin Iris Wang & Yue Zhang------June 1, 2010

Explore Higher-Order Features (1)   Dependency

Parsing by Belief Propagation (Smith &

Eisner, 08)    

Has a first order baseline parser Using a BP network to incorporate higher order features into this first order parser approximately

  Integration

of graph-based and transition-based

models (Zhang & Clark, 08)  

Approximation by beam-search

4

NAACL 2010 Tutorial ------ Qin Iris Wang & Yue Zhang------June 1, 2010

Explore Higher-Order Features (2)   Concise

Integer Linear Programming Formulations for Dependency Parsing (Martins et al. 09)    

Formulate dependency parsing as a polynomial-sized integer linear program Integer linear programming in NLP tutorial this afternoon

5

NAACL 2010 Tutorial ------ Qin Iris Wang & Yue Zhang------June 1, 2010

Use Extra Information Source –Raw Data   Improving

dependency parsing with subtrees from auto-parsed data (W. Chen et al. 09)    

Using a base parser to pare large scale unannotated data Extract subtrees from the auto-parsed data

  Simple

semi-supervised dependency parsing (Koo et al. 08)

  Semi-supervised

convex dependency parsing

(Wang et al. 08) 6

NAACL 2010 Tutorial ------ Qin Iris Wang & Yue Zhang------June 1, 2010

Use Extra Information Source – Bilingual Data   Bilingually-constrained

monolingual shift-reduce

parsing (Huang et al. 09)    

A novel parsing paradigm that is much simpler than bi-parsing Enhance a shift-reduce dependency parser with alignment features to resolve shift-reduce conflicts

7

NAACL 2010 Tutorial ------ Qin Iris Wang & Yue Zhang------June 1, 2010

Use Extra Information Source – Linguistic Rules   Semi-supervised

Learning of Dependency Parsers using Generalized Expectation Criteria (Druck et al. 09)    

Directly use linguistic prior knowledge as a training signal Model parameters are estimated using a generalized expectation (GE) objective function that penalizes the mismatch between model predictions and linguistic expectation constraints.

8

NAACL 2010 Tutorial ------ Qin Iris Wang & Yue Zhang------June 1, 2010

Better Parsing Strategies   Non-projective  

shift-reduce parsing (Nivre, 09)

Expected linear time

  Easy-First

Non-Directional Dependency Parsing

(Goldberg and Elhadad, 10)      

 

Inspired by Shen et al. 07 Use an easy-first order instead, O(nlogn) complexity Allows using more context at each decision

Dynamic programming for incremental parsing (Huang & Sagae, 10)   Linear time NAACL 2010 Tutorial ------ Qin Iris Wang & Yue Zhang------June 1, 2010

9

References  

 

 

 

 

 

 

W. Chen, J. Kazama, K. Uchimoto and K. Torisawa. 2009. Improving Dependency Parsing with Subtrees from Auto-Parsed Data. In Proc. EMNLP. Gregory Druck; Gideon Mann; Andrew McCallum. 2009. Semi-supervised Learning of Dependency Parsers using Generalized Expectation Criteria. In Proc. ACL. Y. Goldberg and M. Elhadad. 2010. An Efficient Algorithm for Easy-First Non-Directional Dependency Parsing. In Proc. NAACL. L. Huang, W. Jiang, and Q. Liu. 2009. Bilingually-Constrained (Monolingual) Shift-Reduce Parsing. In Proc. EMNLP. L. Huang and K. Sagae. 2010. Dynamic Programming for Linear-time Incremental Parsing. In Proc. ACL. W. Jiang and Q. Liu. 2010. Dependency Parsing and Projection Based on Word-Pair Classification. In Proc. ACL. M. Kuhlmann and G. Satta. 2009. Treebank Grammar Techniques for Non-Projective Dependency Parsing. In Proc. EACL.

NAACL 2010 Tutorial ------ Qin Iris Wang & Yue Zhang------June 1, 2010

10

References  

 

 

 

 

A. Martins, N. Smith and E. Xing. 2009. Concise Integer Linear Programming Formulations for Dependency Parsing. In Proc. ACLIJCNLP. J. Nivre and R. McDonald. 2008. Integrating graph-based and transitionbased dependency parsers. In Proc. ACL-HLT. S. Riedel and J. Clarke. 2006. Incremental integer linear programming for non-projective dependency parsing. In Proc. EMNLP. L. Shen, G. Satta and A. K. Joshi. 2007. Guided learning for bidirectional sequence classifica- tion. In Proc. ACL. D. Smith and J. Eisner. 2008. Dependency Parsing by Belief Propagation. In Proc. EMNLP.

11

NAACL 2010 Tutorial ------ Qin Iris Wang & Yue Zhang------June 1, 2010

12

NAACL 2010 Tutorial ------ Qin Iris Wang & Yue Zhang------June 1, 2010

Recent Advances in Dependency Parsing

Jun 1, 2010 - auto-parsed data (W. Chen et al. 09) ... Extract subtrees from the auto-parsed data ... Directly use linguistic prior knowledge as a training signal.

244KB Sizes 4 Downloads 217 Views

Recommend Documents

Experiments in Indian Language Dependency Parsing - web.iiit.ac.in
Language Technologies Research Centre,. International Institute of Information Technology,. Hyderabad, India ... specific to either one particular or all the Indian.

Recent advances in ketene chemistry - Arkivoc
method and used to interpret the excited state of the molecule.16 Electron scattering ... Ketene formation from ground-state oxygen atom reaction with ethylene.

Recent Advances in Nutritional Sciences
men. Subjects consumed a meal containing 15% of energy from protein, 55% from carbohydrate and 30% from fat, in the form of corn oil (CO) and animal fat or MCT oil (56% octanoate, 40% decanoate) in random order. Energy expendi- ture measurements were

Recent Advances in Surfactant EOR.pdf
There was a problem previewing this document. Retrying... Download. Connect more apps... Try one of the apps below to open or edit this item. Recent Advances in Surfactant EOR.pdf. Recent Advances in Surfactant EOR.pdf. Open. Extract. Open with. Sign

Unsupervised Dependency Parsing without ... - Stanford NLP Group
inating the advantage that human annotation has over unsupervised ... of several drawbacks of this practice is that it weak- ens any conclusions that ..... 5http://nlp.stanford.edu/software/ .... off-the-shelf component for tagging-related work.11.

Incremental Joint POS Tagging and Dependency Parsing in Chinese
range syntactic information. Also, the traditional pipeline approach to POS tagging and depen- dency parsing may suffer from the problem of error propagation.

Posterior Sparsity in Unsupervised Dependency Parsing - Journal of ...
Department of Computer and Information Science ... We investigate unsupervised learning methods for dependency parsing models that .... this interpretation best elucidates how the posterior regularization method we propose in Section 4.

Universal Dependency Annotation for Multilingual Parsing
of the Workshop on Treebanks and Linguistic Theo- ries. Sabine Buchholz and Erwin Marsi. 2006. CoNLL-X shared task on multilingual dependency parsing. In.

Corrective Dependency Parsing - Research at Google
dates based on parses generated by an automatic parser. We chose to ..... this task, we experimented with the effect of each feature class being added to the .... Corrective modeling is an approach to repair the output from a system where more.

Tree Revision Learning for Dependency Parsing
Revision learning is performed with a discriminative classi- fier. The revision stage has linear com- plexity and preserves the efficiency of the base parser. We present empirical ... A dependency parse tree encodes useful semantic in- formation for

Posterior Sparsity in Unsupervised Dependency Parsing - Journal of ...
39.2. 4. BS. Ad-Hoc @45. DMV. 55.1. 44.4. 39.4. 5. LsM. Ad-Hoc @15. DMV. 56.2. 48.2. 44.1. 6. LP. Hybrid @45. DMV. 57.1. 48.7. 45.0. Smoothing effects. 7.