| 规则与统计结合分析汉语 |
| |
| 引用本文： | 刘颖.规则与统计结合分析汉语[J].计算机工程与应用,2002,38(7):3-6. |
| |
| 作者姓名： | 刘颖 |
| |
| 作者单位： | 清华大学中文系,北京,100084 |
| |
| 基金项目： | 国家自然科学基金资助项目“基于统计和语义分析的中英文自动文摘的研究”(编号:69972025) |
| |
| 摘 要： | 在自然语言处理中,规则方法和统计方法各有优缺点。采用规则方法进行汉语切分、标注,并采用规则方法与切分标注评分结合进行消歧。对切分标注正确的句子进行句法分析,并采用规则方法与句法语义评分结合进行消歧。根据对多个结果的评分,选择出评分较大的结果,尽可能早地删除掉不正确的结果,从而加快汉语分析的速度,减少分析的空间消耗,提高分析的正确率。
|
| 关 键 词： | 切分 标注 句法分析 规则方法 统计方法 |
| 文章编号： | 1002-8331-(2002)07-0003-04 |
| 修稿时间： | 2002年1月1日 |
Analyzing Chinese with Rule-based Method Combined with Statistics-based Method |
| |
| Liu Ying.Analyzing Chinese with Rule-based Method Combined with Statistics-based Method[J].Computer Engineering and Applications,2002,38(7):3-6. |
| |
| Authors: | Liu Ying |
| |
| Abstract: | Rule-based method and statistics-based method have advantages and disadvantages respectively in natural language processing.The rule-based segmentation and tagging leave some amount of ambiguity that are resolved using statistical approach -segmentation and parts of speech(POS)tagging score.The sentences of correctly segmented and tagged are analyzed,and rule-based method and syntactical and semantic score are adapted to disambiguate.High score results are selected and wrong results are deleted according to score results.Thus the speed is improved,space is saved,and the accuracy rate of analyzing is improved. |
| |
| Keywords: | segmentation tagging syntactic analysis rule-based method statistics-based method |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
