基于决策树的汉语代词共指消解-北京邮电大学学报.pdfVIP

2006 年 8 月 北 京 邮 电 大 学 学 报 Aug .2006 第 29 卷 第 4期 Journal of Beijing University of Posts and Telecommunications Vol .29 No .4 文章编号 :1007-5321(2006)04-0001-05 基于决策树的汉语代词共指消解 王智强 , 李 蕾, 王 枞 (北京邮电大学 智能科学技术研究中心 ,北京 100876) 摘要 :提出一种统计与规则相结合的决策树算法进行汉语代词共指消解 ,利用规则过滤掉属性冲突的反例 ,一定程 度上弥补了决策树算法忽略属性关联性的缺点 .采用 Chinese Treebank 作为语料进行测试 ,手工标注其中的共指 关系和特征向量 ;首先用规 则过滤 ,然后采用 C4.5 决策树算法 选择先行语 . 实验结果显示 ,消解成功率为 82.59 % ,其中人称代词和指示代词的成功率分别为87.60 % 和75.21 % . 关 键 词 :自然语言理解 ;共指消解 ;汉语代词 ;决策树 ;过滤规则 中图分类号 :TP18 文献标识码 :A Chinese Pronominal Coreference Resolution Based on Decision Tree WANG Zhi-qiang , LI Lei , WANG Cong (Research Center of Intelligent Sciences and Technology ,Beijing University of Posts and Telecommunications ,Beijing 100876 ,China) Abstract :An integrated method based on decision tree for Chinese pronominal coreference is proposed . The basic idea is to some extent that filtering out the negative examples based on rules and could com- pensate the drawback of decision tree that ignoring the relationship between attributes . The perfor- mance of the proposed method is tested on Chinese Treebank .In our experiments , the attributes and coreferences are manually labeled ,and then the rule patterns are utilized to feature vectors following the decision tree of C4 .5 algorithm .The success rate is 82.59 % ,in which the rate of personal pro- nouns and demonstrative pronouns are 87.60 % and 75.21 % respectively . Key words :natural language understanding ;coreference resolution ;Chinese pronoun ;decision tree ; filter rules 有同指关系 ,此时的指代消解称为共指消解(core- 0 引 言 [1]