计算机科学 ›› 2014, Vol. 41 ›› Issue (8): 97-100.doi: 10.11896/j.issn.1002-137X.2014.08.021

• 2013年全国理论计算机科学学术年会 • 上一篇    下一篇

蒙古文依存句法分析

苏向东,高光来,闫学亮   

  1. 内蒙古大学计算机学院 呼和浩特010021;内蒙古大学计算机学院 呼和浩特010021;内蒙古大学计算机学院 呼和浩特010021
  • 出版日期:2018-11-14 发布日期:2018-11-14
  • 基金资助:
    本文受国家自然科学基金资助

Dependency Parsing for Traditional Mongolian

SU Xiang-dong,GAO Guang-lai and YAN Xue-liang   

  • Online:2018-11-14 Published:2018-11-14

摘要: 近年来,依存句法分析逐渐成为自然语言处理领域中的研究热点。然而,蒙古文的依存句法分析尚未得到足够的重视。基于最大生成树模型在蒙古文依存关系树库TMDT上进行了蒙古文依存句法分析的研究。在简要介绍蒙古文的特点和蒙古文依存关系树库TMDT之后,详细讨论了最大生成树模型。为找到该模型在蒙古文依存句法分析中合适的特征,重点通过实验对8种特征及其组合在句法分析中的性能进行了比较。结果显示,Basic Unigram Features、Basic Bi-gram Features以及C-C sibling Features这3种特征的组合性能最佳。本研究为蒙古文依存句法分析奠定了基础。

关键词: 蒙古文,依存句法分析,最大生成树,自然语言处理

Abstract: Dependency parsing has become increasingly popular in natural language processing in recent years.Nevertheless,dependency parsing focused on traditional Mongolian has not attracted much attention.We investigated it with Maximum Spanning Tree (MST) based model on Traditional Mongolian dependency treebank (TMDT).This paper briefly introduced traditional Mongolian along with TMDT,and discussesd the details of MST.Much emphasis was placed on the performance comparisons among eight kinds of features and their combinations in order to find a suitable feature representation.Evaluation result shows that the combination of Basic Unigram Features,Basic Bi-gram Features and C-C Sibling Features obtains the best performance.Our work establishes a baseline for dependency parsing of traditional Mongolian.

Key words: Traditional mongolian,Dependency parsing,Maximum spanning tree,Natural language processing

[1] Hideki H.Semantic Dependency Analysis Method for Japanese Based on Optimum Tree Search Algorithm[J].Transactions of Information Processing Society of Japan,2002,43(3):696-707
[2] Mcdonald R,Crammer K,Pereira F.Online Large-Margin Trai-ning of Dependency Parsers[C]∥Proceedings of the 43rd An-nual Meeting on Association for Computational Linguistics.2005:91-99
[3] Eisner J M.Three New Probabilistic Models for DependencyParsing:an Exploration[C]∥Proceedings of the 16th Conference on Computational Linguistics.1996:340-345
[4] Mcdonald R,Pereira F,Ribarov K,et al.Non-projective Depen-dency Parsing Using Spanning Tree Algorithms[C]∥Procee-dings of the Conference on Human Language Technology and Empirical Methods in Natural Language Processing.2005:523-530
[5] Chu Y-J,Liu T-H.On the Shortest Arborescence of a Directed Graph[J].Science Sinica,1965,14:1396-1400
[6] Jack E.Optimum Branchings[J].Journal of Research of the National Bureau of Standards,1967,71B:233-240
[7] Crammer K,Singer Y.Ultraconservative Online Algorithms for Multiclass Problems[J].Journal of Machine Learning Research,2003,3:951-991
[8] Crammer K,Dekel O,Shalev-Shwartz S,et al.Online Passive-Aggressive Algorithms[C]∥Procings of the Sixteenth Annual Conference on Neural Information Processing Systems (NIPS).2003
[9] Mel′ˇuk I A.Levels of Dependency in Linguistic Description:Concepts and Problems[J].Dependency and Valency,2003(1):188-230
[10] Hudson R.An Introduction to Word Grammar[M].Cambridge:Cambridge University Press,2010
[11] Nivre J.Dependency Grammar and Dependency Parsing[R].School of Mathematics and Systems Engineering,Vxj University,2005
[12] 清格尔泰.蒙古语语法[M].呼和浩特:内蒙古人民出版社,1992
[13] Georgiadis L.Arborescence Optimization Problems Solvable by Edmonds’Algorithm[J].Theor.Comput.Sci.,2003,301(1-3):427-437

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!