权益分类	功能权益	普通用户	{{item.name}}会员
{{category.name}}	{{benefitItem.name}}

Development of the system that punctuates a Chinese character letter line of Japanese historical materials

日本史料汉字字母行标点系统的开发

基本信息

批准号：
06551005
负责人：
NAGAMURA Makoto
金额：
$ 6.72万
依托单位：
JAPAN WPMEN'S UNIVERSITY
依托单位国家：
日本
项目类别：
Grant-in-Aid for Scientific Research (A)
财政年份：
1994
资助国家：
日本
起止时间：
1994 至 1996
项目状态：
已结题

项目摘要

For the research of Japanese history, full text detabase of historical materials are very necessary instruments. However, it is difficult to make index easily from a Chinese writing. A historical material text without an index is the enumeration of a plain Chinese character. Thereupon, we developed such computer processing system that makes an index easily from a Chinese writing. And we examined the policy of system development and would adopt the method called a morpheme analysis. First of all to a letter line of classics Japanese, marks are attached and divide into words and phrases. Words and phrases file are created, by accumulating words and phrases from a sentence in accordance with marks. Application of this file, we are able to develop the system that punctuates automatically words and phrases, by utilizing. In this way, a lot of words and phrases are accumulated out of a letter line of a Chinese writin. Various kinds of indexes are made from this words and phrases file. Expecially the form of the index called KWIC index is very useful for Japanese history research, because it can read contents of historical material directly. Such processing system is the result of this research. However, processing that punctuates the sentence of a Chinese writing correctly by a computer, there is the room for an improvement yet. If the pattern of the connection of words and phrases was confirmed numerously, by moving this processing system, it depends on a grammatical rule and will materialize the punctuation of a more exact sentence.

对于日本史的研究来说，史料的全文整理是非常必要的工具。然而，很难从一个中文写作很容易作出索引。没有索引的史料文本是对一个普通汉字的列举。因此，我们开发了这样的计算机处理系统，使索引容易从一个中文写作。我们研究了系统开发的策略，并将采用称为语素分析的方法。首先对一个字母行的经典日语进行标记，并将其分为单词和短语。单词和短语文件是通过根据标记从句子中积累单词和短语来创建的。应用该文件，我们可以开发自动标点系统，通过利用。这样，一个汉字的一行字就积累了大量的单词和短语。各种索引是从这个单词和短语文件。特别是KWIC索引的形式，可以直接阅读史料的内容，对日本历史研究非常有用。这种处理系统就是本研究的成果。然而，在计算机上对汉语句子进行正确标点的处理，还存在着改进的空间。如果单词和短语的连接模式被大量确认，通过移动这个处理系统，它依赖于语法规则，并将实现更精确句子的标点符号。