Title: | 机器翻译中词典索引的构造--KTI方法及其应用 |
Author: | 傅俊
|
Issued Date: | 1995
|
Major: | 计算机软件(computer software)
|
Degree Grantor: | 中国科学院软件研究所
|
Place of Degree Grantor: | 中国科学院软件研究所
|
Degree Level: | 博士
|
Keyword: | 机器词典
; 汉语自动分词
; 基于模板的机器翻译
; 多语言机器翻译
|
Alternative Title: | The Construction of Dictionary Index in MT:KTI Method and its Application
|
Abstract: | 本文给出了直接访问由可变长记录组成的顺序式词典文件中的各个词条的一种方法。这种方法以Key-tree索引为基础,避免了重新组织文件。Key-tree索引是一个深度可变的多层次索引,它或者它的变形可用以支持大规模词典,并可在其上进行高速查找。本方法可满足查找一个次条只需一次磁盘访问的理想要求。在这种索引之上,很容易建立起多种特殊查找方式,如头匹配的查找,最大、最小匹配的查找等。本文给出了一些特殊查找算法,并介绍了用这些查找算法实现的若干应用。 |
English Abstract: | In this paper, a method to access any one of the unfixed size records in a sequencial dictionary file is described. This method takes takes key-tree as its data structure for the dictionary index, which mades it possible not rebuilding the dictionary file. Key-tree index is multi-level and with unfixed depth. It or its variant forms support large dictionaries and high speed searching on them. With this method used, any record can be found in only one time of disk accessing. Some specialalgorithms, including Head Matching Searching, Maximum Matching Searching, Minimum Matching Searching, etc., can be easily designed and implemented with this kind of index. This paper also shows some applications using these algorithms. |
Language: | 中文
|
Content Type: | 学位论文
|
URI: | http://ir.iscas.ac.cn/handle/311060/6854
|
Appears in Collections: | 中科院软件所
|
File Name/ File Size |
Content Type |
Version |
Access |
License |
|
N91052.pdf(3596KB) | -- | -- | 限制开放 | -- | 联系获取全文 |
|
Recommended Citation: |
傅俊. 机器翻译中词典索引的构造--KTI方法及其应用[D]. 中国科学院软件研究所. 中国科学院软件研究所. 1995-01-01.
|
|
|