泰语文语转换系统中的文本略论措施探讨Analysis of text to text system conversion in Thai 摘要:（摘要内容经过系统自动伪原创处理以避免复制，泰语毕业论文，下载原文正常，内容请直接查看目录。） 泰国位于中国南部,处于西北亚的中间,是我国的友爱邻邦,泰国与中国商业来往频仍,和我国在经济、政治范畴有主要协作关系。泰语是泰国的官方说话,今朝有6000多万生齿运用泰语,泰语是一种剖析型、孤立型说话,它多由单音节词组成根本辞汇。同时,泰语也是一门音调说话,音调可以用来辨别分歧的辞汇和语法。西北亚市场包含着伟大的潜力,西北亚说话将会成为语音分解范畴的研究重点。论文的重要任务包含：].从专业泰语在线辞书下载词条包,遴选出包含经常使用词、复合词、地名、数目词、借词等词条,并经由过程在线辞书人工查询并添加尺度音标和词性信息,构建泰语辞书。2.从泰语专业书本和网站中遴选经常使用语句,并对语料停止进一步挑选,去除长度、格局不适合的语句,最初留下作为前端文天职析的语料。3.在构建好泰语辞书的基本上,停止泰语TTS体系前端文天职析,针对泰语说话特点,设计基于泰语辞书的前向后向最年夜婚配算法停止分词,再将分词成果运用辞书调换成对应泰语音节信息。4.针对基于泰语辞书前向后向最年夜婚配算法未婚配到的词语(即未知词),设计基于泰语音节拼写规矩的处置办法停止处置。5.设计一个改良的泰语罗马化编码计划,在上述进程完成的基本上,用该编码计划对泰语文本停止编码,再与尺度泰语读音停止比拟。本论文经由过程前向后向婚配算法对泰语文本停止基于泰语辞书的分词,曾经可以或许将辞书内包括的词语准确切分出来,对未知词停止处置后,将文本尽可能表示成音节组合情势,经试验研究,初步分词准确率到达78%；语音分解体系须要将文本罗马化以提取音节声调信息,是以还须要将泰语停止罗马化,经由改良的泰语罗马化编码计划可以或许更精确的表达出泰语音节和声调信息。 Abstract: Thailand is located in southern China, and in the middle of Southeast Asia is China's friendly neighbor, Thailand and China commercial intercourse is frequent, and China in economic and political areas is the main collaborative relationship. Thai is Thailand's official language. There are currently 60 million population of Thai, Thai is talking to an analysis of the type, isolated type, it from monosyllabic words composed of basic vocabulary. At the same time, Thai is a voice tone, can be used to identify differences in vocabulary and grammar. Southeast Asia market contains great potential, Southeast Asia will become the research focus to the category of speech decomposition. Main task of this paper includes:. From a professional Thai online dictionaries download entry package, the selection contains frequently used words, compound words, place names, the number of words, words such as entry, and through the process of online dictionaries manual query and add scale phonetic and grammatical information, constructing Thai phrase book. 2. From the Thai professional books and websites selection often use the statement, and the corpus to stop further selection, length of removal, pattern is not suitable for the statement was initially left behind as a front-end Wentian post analysis of the corpus. In the construction of good Thai dictionary basically, stop Thai TTS system front-end Wentian job analysis, for Thai speech characteristic, design based on Thai dictionary before back to the most of the eve of the matching algorithm stop word segmentation, the segmentation results of the application dictionary change into corresponding Thai syllable information. 4. For Thai dictionary based on backward maximum matching algorithm to single to (i.e., unknown words) words, design based on Thai syllable spelling rules of solution to stop the disposal of. 5. Design a modified Thai Rome of coding schemes, based on the completion of the process, the coding scheme of Thai text to stop encoding, and Thai pronunciation scale compared. In this paper, through the process of backward matching algorithm of Thai text stop Thai dictionary based segmentation, once can perhaps the dictionary include accurate word segmentation, of unknown words to stop the disposal of the text as much as possible into syllables situation, the experimental research, the initial segmentation accuracy rate reached 78%; speech synthesis system needs to be text Romanized to extract the syllable and the tone information is also need to Thai stop Roman, through improved Thai Roman encoding scheme can perhaps more accurate express Thai syllable and tone information. 目录: 摘要 3-4 Abstract 4-5 第一章 绪论 8-13 1.1 引言 8-10 1.2 语音合成技术的发展历史 10-11 1.3 文本略论处理在语音合成中的重要性 11 1.4 探讨思路以及论文的主要工作 11-12 1.5 论文组织 12-13 第二章 泰语文语转换系统 13-23 2.1 泰语简介 13-15 2.2 探讨近况 15-16 2.3 文语转换的关键技术 16-20 2.3.1 TTS系统概述 16-17 2.3.2 文语转换系统中的语音波形合成措施 17-20 2.4 构建泰语文语转换系统的过程 20-22 2.5 本章小结 22-23 第三章 泰语文本略论措施 23-35 3.1 文本略论概述 23-24 3.2 分词 24-25 3.2.1 分词技术概述 24-25 3.3 实验方案 25-32 3.3.1 构建泰语词典 25-27 3.3.2 泰语自动分词处理 27-31 3.3.3 未知词处理 31-32 3.4 实验结果与略论 32-34 3.4.1 算法复杂度略论 32-33 3.4.2 泰语分词结果略论 33-34 3.5 本章总结 34-35 第四章 泰语文本的罗马化 35-48 4.1 罗马化概述及背景 35-36 4.1.1 音译 35-36 4.1.2 转录 36 4.2 罗马化的运用 36 4.3 泰语罗马化编码标准 36-41 4.4 实验方案 41-46 4.4.1 改进的泰语罗马化编码方案 41-44 4.4.2 音调修改方案 44-45 4.4.3 特殊声调处理方案 45-46 4.4.4 特殊符号处理 46 4.5 实验结果与略论 46-47 4.6 本章总结 47-48 第五章 总结与展望 48-51 5.1 总结 48-49 5.2 展望 49-51 附录 51-53 参考文献 53-57 致谢 57 |
