应用二叉树剪枝识别韵律短语边界

摘要
参考文献
相关文章 (15)

全文: PDF (314 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要句子的韵律短语识别是语音合成的重要研究内容。本文提出了应用统计语言模型生成的二叉树,结合最大熵方法识别待合成汉语句子的语音停顿点。文中给出了二叉树相关的模型训练和生成算法;二叉树与语音停顿点之间的关系;在最大熵方法中应用二叉树剪枝识别句子的韵律短语。实验结果表明,在搜索算法中,利用二叉树进行剪枝,可以很大程度上提高语音停顿预测的正确率和召回率,基于试验数据的f-Score提高了近35%。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	荀恩东
	钱揖丽
	郭庆
	宋柔

关键词 ：人工智能, 自然语言处理, 统计语言模型, 二叉树, 韵律短语, 最大熵

Abstract：It is important to recognize the prosodic phrase breaks in text-to-speech. In this paper, a new method is introduced for this purpose,which uses binary tree as pruning strategy in the Maximal Entropy Model (MaxEnt) framework. First of all, the concep t of binary tree generated from a statistical language model is given. Then the process of generating the binary tree is discussed. In the process of applying MaxEnt to seeking optimal prosodic phrases, the binary tree is exploited so as to narrow the search space and improve the performance. Experimental results show that the F-score of predicating prosodic phrase breaks is about 35% better than the previous system, in which the binary tree strategy is not adopted.

Key words： artificial intelligence natural language processing statistical language model binary tree prosodic phrase Maximal Entropy Model

收稿日期: 2005-07-20

中图分类号:

TP391

基金资助:国家自然科学基金资助项目(60573184)

作者简介: 荀恩东(1967—),男,博士,副教授,主要研究方向为自然语言处理.

链接本文:

http://jcip.cipsc.org.cn/CN/Y2006/V20/I3/3