汉语句法树库一致性检验方法研究  

Study of Keeping Consistency of Chinese Corpus of Complete Parsing

作　　者：魏莉[1] 谭红叶[1] 郑家恒[1] 孙健[2] 

机构地区：[1]山西大学计算机与信息技术学院,山西太原030006 [2]阿里巴巴集团研发院,北京130000

出　　处：《广西师范大学学报（自然科学版）》2010年第1期139-142,共4页Journal of Guangxi Normal University:Natural Science Edition

基　　金：国家自然科学基金资助项目(60775041)

摘　　要：为了进一步提高完全句法分析标注的准确率,对人工修正的完全句法分析语料进行剖析,从分词、词性和句法结构三个层面检验一致性,总结标注结果不一致的类型,并提出基于分层的自动发现不一致现象的方法及相应的消解策略。实验表明,利用该方法可使语料库标注的准确率提高2.5%。In order to improve the accuracy of the corpus of complete parsing,this paper analyzes the corpus which was amended by people and summarizes the reason leading to the annotation inconsistency. Moreover,some strategies to eliminate inconsistencies and explored disambiguation type are proposed from various hierarchies such as the word segmentation,Pos tagging,and parsing structure. Experiments show that the method described in this article can improve the accuracy of corpus annotation by 2. 5%.

关 键 词：中文信息处理 语料库 完全句法分析 一致性 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]