电子邮件文档中维、英、汉文种识别技术研究与实现-软件工程专业论文.docxVIP

摘 要 近年来，在新疆互联网应用本地化过程中随着技术的不步进步和发展，维吾 尔文新老文字在信息处理中得到了广泛使用，特别是在电子邮件中尤为突出。对 于该文种的识别，具有非常鲜明的新疆地域特色，需要研究开发相应识别技术和 方法。 本文针对在工作中实际急需解决邮件中维英汉等文字识别的问题,在详细分析 了电子邮件结构、汉字编码识别技术的基础上，提出了字典驱动的电子邮件文档 语种识别模型，采用新维文语料库和字典构造、字典查询内存快速定位等方式方 法，重点解决了对维吾尔新文字和老文字的识别技术难题。开发了专门的电子邮 件文档维英汉等文字文种识别应用系统，提高系统对海量邮件识别的运行效率。 该系统在通过大量测试、分析和不断完善后，目前已在实际业务中得以应用， 各项功能、指标达到系统设计要求，解决了实际工作中迫切需要解决的技术问题， 取得了较好的效果。 关键词：电子邮件 维英汉文种识别 字典驱动 内存快速定位 Abstract In recent years, as the technology developed continuously in the localization of network application, both the new and old Uyghur characters are widely used in information processing, especially in e-mails. As the recognition of Uyghur characters has typical characteristic of Xinjiang, new research, new distinguishing technique and method are needed. In view of the problems that need to be solved in practical work, like the recognition of Uyghur, Han language and English, after a close analysis of e-mail structure and the technique of distinguishing the coding of Chinese characters, the author puts forward a dictionary-driven model for language distinguishing in e-mails. It focus on solving the difficulties in distinguishing between new and old Uyghur characters and develops a special kind of e-mail text recognition applications by means of establishing corpus of new Uyghur, dictionary construction and quick orientation of memory and so on. After abundant experiments and analysis as well as constinuous optimization, now this system has been used in practical work in which its functions and performance meet the requirements of system design and solved technical issues which are in urgent need to be solved, and attained certain achievements. Keyword ： E-mails recognition of Uyghur Chinese and English language dictionary-driven quick orientation of memory 目 录 第一章 绪论1 1.1 论文的研究背景及选题1 1.2 国内外研究现状与分析1 1.3 论文研究的主要内容3 1.4 论文各章节的安排3 第二章 相关技术概述5 HYPERLINK \l _TOC_250001 2.1 RFC822 邮件格式 5 HYPERLINK \l _TOC_250000 2.2 MIME 协议邮件格式 6 MIME 邮件的组织结构 7