普通话智能测试系统的语音识别网络研究

现行的计算机辅助普通话水平测试系统主要采用隐马尔科夫模型的对数后验概率算法来衡量考生的发音质量，但是HMM模型之间易混淆。为提高系统测试的效度和信度，将普通话发音中的语言学知识引入测试系统，重构算法的识别网络，对算法的概率空间进行优化。实验结果表明，改进后的识别网络能够显著缩短系统的运算时间，有效降低概率空间对评分的影响，提高系统的评测性能。

Research on Speech Recognition Network in Putonghua Level Test System

The existing computer-aided system uses the algorithm of HMM based log posterior probability to judge the tester＆#39;s pro-nunciation , but the confusion between HMM models is big .In order to improve the validity and reliability of the system , the author re-constructs the recognition network in algorithm based on the introduction of linguistic knowledge of Putonghua pronunciation , and opti-mizes the probability spaces in algorithm .Experimental results indicate that the improved recognition networks can not only significantly reduce the system＆#39;s operation time, but also effectively reduce the probability space impact on scoring , and improve the system of eval-uating performance .