基于web的藏文文本自动分类研究与实现

发布时间：2018-04-05 17:40

本文选题：藏文网页　切入点：自动分类　出处：《青海民族大学》2012年硕士论文

【摘要】：随着互联网的普及和藏文信息技术的发展，藏文网站的数量越来越多，藏文信息资源从匮乏的时代过渡到极为丰富的藏文数字化时代。但面对海量的藏文信息资源，人们又很难得到实际需要的信息。因此，，人们常对藏文网页按照其内容进行分类。目前，用传统的分类方法来处理是不切实际的，必须借助计算机对藏文网页进行自动分类，一来减少了人力的耗费和财力的投资，二来提高了网页分类的效率和准确率。因此，用计算机进行藏文网页文本的自动分类，可以满足人们所期望对各类藏文文本分类应用需求，这也成为迫在眉睫的研究课题。笔者认为对藏文网页文本自动分类的研究在藏文搜索引擎、藏文数字图书馆和藏文出版等领域，以及对推动藏文信息技术的发展与应用有着重要的现实意义和使用价值。本文主要对藏文网页的净化、藏文分词、特征选取、权值计算、分类算法等相关技术进行了较深入的分析和研究，其中重点研究了藏文分词和藏文网页分类算法，并借鉴现有的中英文分类算法，提出了符合藏文文法及藏文字结构特征的分类算法，即词频统计+词匹配法、基于标题的藏文网页分类算法和基于导航栏的藏文网页分类算法的一种新的混合分类算法，该方法大大提高了基于web的藏文文本自动分类系统的查全率和查准率。最后实现了基于web的藏文文本自动分类系统，同时对实验结果进行了分析并具有良好的分类结果。
[Abstract]:With the popularization of the Internet and the development of Tibetan information technology, the number of Tibetan websites is increasing, and the Tibetan information resources transition from the era of scarcity to the extremely rich digital era of Tibetan language.But in the face of massive Tibetan information resources, it is difficult for people to get the actual information.Therefore, people often classify Tibetan web pages according to their content.At present, it is impractical to use the traditional classification method to classify Tibetan web pages automatically with the help of computer, which not only reduces the cost of manpower and financial resources, but also improves the efficiency and accuracy of web page classification.Therefore, the automatic classification of Tibetan web pages by computer can meet the needs of all kinds of Tibetan text classification, which has become an urgent research topic.The author thinks that the research on automatic text classification of Tibetan web pages is of great practical significance and practical value in Tibetan search engine, Tibetan digital library and Tibetan publishing, as well as in promoting the development and application of Tibetan information technology.In this paper, the purification of Tibetan web pages, Tibetan word segmentation, feature selection, weight calculation, classification algorithm and other related techniques are analyzed and studied deeply, among which, Tibetan word segmentation and Tibetan web page classification algorithms are mainly studied.Using the existing Chinese and English classification algorithms for reference, this paper puts forward a classification algorithm that accords with Tibetan grammar and Tibetan characters' structural features, that is, word frequency statistical word matching method.A new hybrid classification algorithm for Tibetan web pages based on title and navigation bar is proposed, which greatly improves the recall and precision of the automatic classification system of Tibetan text based on web.Finally, an automatic Tibetan text classification system based on web is implemented, and the experimental results are analyzed and good classification results are obtained.
【学位授予单位】：青海民族大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：H214

【相似文献】