| 基于TBL算法的汉语韵律词预测 |
| |
| 引用本文： | 陈龙,杨鸿武,蔡莲红.基于TBL算法的汉语韵律词预测[J].西北师范大学学报,2008,44(1):47-51. |
| |
| 作者姓名： | 陈龙 杨鸿武 蔡莲红 |
| |
| 作者单位： | [1]西北师范大学物理与电子工程学院,甘肃兰州730070 [2]清华大学计算机科学与技术系,北京100084 |
| |
| 基金项目： | 西北师范大学科研骨干培育项目 |
| |
| 摘 要： | 提出了一种新的汉语韵律词预测方法.利用标注过的语料,分析了语法词与韵律词之间的关系,发现24%的韵律词由不同语法词组合而成,语法词的词长是确定韵律词边界的主要特征.基于以上分析,实现了一种基于错误驱动的规则学习算法(TBL)的韵律词预测方法.实验结果表明,所提出的方法在测试集上能够达到97.5%的预测精度.
|
| 关 键 词： | 韵律词 语法词 TBL算法 文语转换 规则学习算法 汉语韵律词 预测精度 learning based word chinese 测试集 预测方法 结果 实验 错误驱动 特征 边界 词长 词组合 同语 发现 关系 语法词 |
| 文章编号： | 1001-988X（2008）01-0047-05 |
| 收稿时间： | 2007-06-22 |
| 修稿时间： | 2007-10-23 |
Predicting chinese prosodic word based on transformation-based error-driven learning |
| |
| CHEN Long,YANG Hong-wu,CAI Lian-hong.Predicting chinese prosodic word based on transformation-based error-driven learning[J].Journal of Northwest Normal University Natural Science (Bimonthly),2008,44(1):47-51. |
| |
| Authors: | CHEN Long YANG Hong-wu CAI Lian-hong |
| |
| Abstract: | A novel approach for predicting chinese prosodic word is introduced. By analyzing a manual tagged corpus, the relationship between lexical word and prosodic word are found. The analysis results show that 24% prosodic words consist of two or more lexical words, and the length of lexical word is a most important feature for predicting prosodic words. A transformation-based error-driven learning algorithm is proposed to predicting prosodic word with lexical features. Experiments demonstrat that the proposed approach outperform other methods with over 97.5% predicting precision. |
| |
| Keywords: | prosodic word lexical word transformation-based error-driven learning text to speech |
| 本文献已被 维普 万方数据 等数据库收录！ |
