作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2011, Vol. 37 ›› Issue (23): 20-23. doi: 10.3969/j.issn.1000-3428.2011.23.007

• 博士论文 • 上一篇    下一篇

基于互补模型的汉语重音检测

倪崇嘉1,2,刘文举2,徐 波2   

  1. (1. 山东财政学院统计与数理学院,济南 250014;2. 中国科学院自动化研究所模式识别国家重点实验室,北京 100190)
  • 收稿日期:2011-05-17 出版日期:2011-12-05 发布日期:2011-12-05
  • 作者简介:倪崇嘉(1979-),男,讲师、博士,主研方向:语音识别,韵律模型;刘文举、徐 波,研究员、博士、博士生导师
  • 基金资助:
    国家自然科学基金资助项目(90820303, 60675026, 9082 0011);国家“863”计划基金资助项目(20060101Z4073, 2006AA01Z 194);国家“973”计划基金资助项目(2004CB318105)

Mandarin Stress Detection Based on Complementary Model

NI Chong-jia 1,2, LIU Wen-ju 2, XU Bo 2   

  1. (1. School of Statistics and Mathematics, Shandong University of Finance, Jinan 250014, China; 2. National Key Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China)
  • Received:2011-05-17 Online:2011-12-05 Published:2011-12-05

摘要: 针对现有汉语重音检测方法正确率较低的问题,利用声学、词典和语法相关特征的不同分类器组合,基于Boosting分类回归树+条件随机场的互补模型,提出一种改进的汉语重音检测方法。在ASCCD语料库上的实验结果表明,该方法能获得84.9%的重音检测正确率,相比基于神经网络+决策树的基线系统提高2.7%。

关键词: 重音, 互补模型, Boosting分类回归树, 条件随机场, 神经网络, 支持向量机

Abstract: Aiming at the problem that existing mandarin stress detection method has low accuracy rate, this paper proposes the complementary model method based on Boosting Classification and Regression Tree(CART) and Conditional Random Fields(CRFs). It is the combination of different classifiers to detect Mandarin character stress by using the acoustic, lexical and syntactic related features. In the ASCCD corpus, the complementary model achieves 84.9% stress detection correct rate, and there is 2.7% improvement when compared with the baseline system based on Neural Network(NN) and decision tree.

Key words: stress, complementary model, Boosting Classification and Regression Tree(CART), Conditional Random Fields(CRFs), Neural Network(NN), Support Vector Machine(SVM)

中图分类号: