Nowadays, computer-assisted language learning system is more and more popular with modern human beings. The pronunciation quality assessment system is the core module in the learning system. As the most effective confidence measure in assessment system, the posterior probability is used widely, in which some tricks are applied to reduce the computation complexity.  

However, this traditional algorithm has some defects and needs to be improved. First, the traditional algorithm adopts the method of maximum instead of sum in the calculation of the denominator, which seriously reduces the accuracy of posterior probability. Second, the posterior probability is normalized by its segment time. In fact, the acoustic likelihood is more related with time and grows with the frame number.

So aiming at solving the two problems above, researchers of Institute of Acoustics, Chinese Academy of Sciences carried out a series of experiments and proposed the improvement method. Taking into account both computation complexity and system performance, they processes an algorithm based on phoneme confusion extended network. To the second defect, they proposed the acoustic likelihood based normalization algorithm. The experimental results show that compared to traditional algorithm, the proposed algorithm can improve system performance significantly, about 35% average score error rate relatively, and the computation complexity does not increased.