一种基于规则的无监督词性标注方法  

Rule-Based Method for Unsupervised Part-of-Speech Tagging

作　　者：彭涛[1] 戴耀康[1] 朱枫彤 张邦佐[2] 刘露[1] 闫昭[1] 钱锋[1] 

机构地区：[1]吉林大学计算机科学与技术学院,长春130012 [2]东北师范大学计算机科学与信息技术学院,长春130117

出　　处：《吉林大学学报（理学版）》2015年第5期956-962,共7页Journal of Jilin University:Science Edition

基　　金：国家自然科学基金(批准号:71473035);教育部人文社会科学研究项目(批准号:14YJA870010);吉林省工业技术研究和开发项目(批准号:JF2012c016-2);吉林省科技重点攻关项目(批准号:20150204040GX);吉林大学研究生创新基金(批准号:2015040)

摘　　要：提出一种基于规则的无监督词性标注方法,利用200多条英语语法规则,创建26个规则函数,先将输入的待标注英语句子进行预处理后得到初始标记,再对每个单词调用规则函数,最终得到标注后的英语句子.通过对Brown语料库的实验,词性标注的正确率达到93.95%.实验结果表明,本文方法可行、有效,能很好地提高英语词性标注的准确率.A rule-based tagging method for unsupervised part-of-speech was proposed.More than 200 grammar rules were used to create 26 kinds of rules functions.After it was preprocessed,the initial tags of words in the input sentence were obtained,the 26 kinds of rules functions were applied to each word to attain all the tags of the input sentence.Experimental results on Brown corpus show that the accuracy of our method is up to 93.95%,thus,our rule-based method is feasible and effective,and improves the accuracy and the simplicity of English part-of-speech tagging.

关 键 词：词性标注 基于规则 无监督学习 规则函数 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]