Probabilistic Parsing Action Models for Multi-Lingual Dependency Parsing

Proceedings of the CoNLL Shared Task Session of EMNLP-CoNLL 2007, pp. 940–946, Prague, June 2007. c⃝2007 Association for Computational Linguistics Probabilistic Parsing Action Models for Multi-lingual Dependency Parsing Xiangyu Duan Institute of Automation, Chi- nese Academy of Sciences xyduan@nlpr.ia.ac.cn Jun Zhao Institute of Automation, Chi- nese Academy of Sciences jzhao@nlpr.ia.ac.cn Bo Xu Institute of Automation, Chi- nese Academy of Sciences xubo@hitic.ia.ac.cn Abstract Deterministic dependency parsers use pars- ing actions to construct dependencies. These parsers do not compute the probabil- ity of the whole dependency tree. They only determine parsing actions stepwisely by a trained classifier. To globally model parsing actions of all steps that are taken on the input sentence, we propose two kinds of probabilistic parsing action models that can compute the probability of the whole dependency tree. The tree with the maxi- mal probability is outputted. The experi- ments are carried on 10 languages, and the results show that our probabilistic parsing action models outperform the original de- terministic dependency parser. 1 Introduction The target of CoNLL 2007 shared task (Nivre et al., 2007) is to parse texts in multiple languages by using a single dependency parser that has the ca- pacity to learn from treebank data. Among parsers participating in CoNLL 2006 shared task (Buchholz et al., 2006), deterministic dependency parser shows great efficiency in time and compa- rable performances for multi-lingual dependency parsing (Nivre et al., 2006). Deterministic parser regards parsing as a sequence of parsing actions that are taken step by step on the input sentence. Parsing actions construct dependency relations be- tween words. Deterministic dependency parser does not score the entire dependency tree as most of state-of-the- art parsers. They only stepwisely choose the most probable parsing action. In this paper, to globally model parsing actions of all steps that are taken on the input sentence, we propose two kinds of prob- abilistic parsing action models that can compute the entire dependency tree’s probability. Experi- ments are evaluated on diverse data set of 10 lan- guages provided by CoNLL 2007 shared-task (Nivre et al., 2007). Results show that our prob- abilistic parsing action models outperform the original deterministic dependency parser. We also present a general error analysis across a wide set of languages plus a detailed error analysis of Chinese. Next we briefly introduce the original determi- nistic dependency parsing algorithm that is a basic component of our models. 2 Introduction of Deterministic Depend- ency Parsing There are mainly two representative deterministic dependency parsing algorithms proposed respec- tively by Nivre (2003), Yamada and Matsumoto (2003). Here we briefly introduce Yamada and Matsumoto’s algorithm, which is adopted by our models, to illustrate deterministic dependency parsing. The other representative method of Nivre also parses sentences in a similar deterministic manner except different data structure and parsing actions. Yamada’s method originally focuses on unla- beled dependency parsing. Three kinds of parsing actions are applied to construct the dependency between two focus words. The two focus words are the current sub tree’s root and the succeeding (right) sub …

Content truncated to 500 words