基于实例的中文分词-词性标注方法的应用研究  

Application Study of Example Based Chinese Word Segmentation and Part-of-speech Tagging Method

作　　者：姜涛[1] 姚天顺[1] 张俐[1] 

机构地区：[1]东北大学自然语言处理实验室,辽宁沈阳110004

出　　处：《小型微型计算机系统》2007年第11期2090-2093,共4页Journal of Chinese Computer Systems

摘　　要：通过实验证明基于实例的中文分词-词性标注(下文简称为EBST,Example-Based Chinese word Segment and Tagging)系统对训练语料相关的文本具有非常好的标注性能.实验结果显示了EBST系统的分词-词性标注不仅具有非常高的准确率,而且和训练语料的标注保持了很好的一致性.这使得EBST系统非常适合于在基于实例的机器翻译(Example-Based Machine Translation,EBMT)系统中的应用.本文给出了EBST在EBMT系统中的应用实例及相应的实验结果.The experiments in this paper proved that the example-base Chinese word segmentation and part-of-speech tagging （EBST for abbr. ） system has very good performance on training corpus-relevant texts. The experiments show that the results of segmentation and part-of-speech given by the experiment EBST system have not only very high precision but also very good coherence with the training corpus. This makes EBST systems very suitable for the applications within example-base machine translation systems. An example of this kind application is also given in this paper ,as well as its experiment results.

关 键 词：中文分词-词性标注 语料库 EBMT 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]