蒙汉英多语种词典资源管理平台的设计与实现

发布时间：2019-04-04 12:21

【摘要】：本文以传统蒙古文、西里尔蒙古文、汉文和英文的多语种对照资源库为对象,搭建了蒙汉英多语种词典资源管理平台。平台的资源库涵盖了目前主流的传统蒙古文-汉文、西里尔蒙古文-汉文、传统蒙古文-英文和西里尔蒙古文-英文等30多个词典,对照词总条数达到近150多万。该多语种资源库的建立为蒙古文自然语言处理提供了基础资源,并对蒙古语言文字的保护与传承具有重大意义。本资源管理平台主要有根据不同语种查询词条解释、资源库的综合管理与维护功能,使得该资源库的管理和使用更加方便、准确。因为对照词库的词条数量较多,所以本文对资源库建立了索引来提升检索的效率。本资源管理平台针对于查询功能模块,采用了一种B+树的数据结构来对传统蒙古文-汉文、西里尔蒙古文-汉文、传统蒙古文-英文和西里尔蒙古文-英文等词典建立索引。B+树索引是数据库中常用的一种索引技术,B+树索引支持顺序查找和随机查找,因此通过B+树的索引算法有效提高了词典查询效率。由于蒙古文属于黏着语,所以本文根据蒙古文的构词特点对传统蒙古文进行了词干后缀切分处理,使得传统蒙古文的查询更加准确有效。系统中传统蒙古文的默认排序功能与平时的蒙古文字母排序顺序有所区别,因此本文建立了基于规则的传统蒙古文排序系统,并应用在传统蒙古文的检索结果和资源管理维护中。
[Abstract]:The resource management platform of Mongolian-English multilingual dictionary is set up in this paper, which is based on the multi-language control resource base of traditional Mongolian, Cyrmund, Chinese and English. The resource bank of the platform covers more than 30 dictionaries, such as traditional Mongolian-Chinese, Cyrillic Mongolian-Chinese, traditional Mongolian-English and Cyrillic Mongolian-English. The total number of control words is more than 1.5 million. The establishment of this multilingual resource database provides the basic resources for Mongolian natural language processing and has great significance for the protection and inheritance of Mongolian language and characters. This resource management platform mainly has the function of comprehensive management and maintenance according to different languages, which makes the management and use of the resource database more convenient and accurate. Because of the large number of entries in the reference database, the index of the database is established to improve the efficiency of retrieval. In view of the query function module, this resource management platform adopts a B-tree data structure to the traditional Mongolian-Chinese, Cyrillic-Chinese, Traditional Mongolian-English and Cyrillic-English dictionaries are indexed. B-tree index is a commonly used index technology in database. B-tree index supports sequential search and random search. Therefore, the efficiency of dictionary query is improved effectively by the index algorithm of B-tree. Because Mongolian belongs to adhesive language, this paper deals with the stem suffix segmentation of the traditional Mongolian according to the characteristics of Mongolian word formation, which makes the traditional Mongolian query more accurate and effective. The default sort function of traditional Mongolian in the system is different from the order of Mongolian alphabetical order, so the traditional Mongolian sorting system based on rules is established in this paper. And applied in the traditional Mongolian retrieval results and resource management and maintenance.
【学位授予单位】：内蒙古大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP311.52

【相似文献】