Automatic Grading System for Mandarin Proficiency Test based on PSO-ANN

The National Mandarin Proficiency Test(NMPT) to be completed by computer-aided evaluation, not only can remove the human factor to establish uniform standards, but also save time, improve efficiency, accelerate the promotion and popularization of mandarin. In this paper, we focus on the optimization of artificial neural network by PSO for automatic grading the NMPT. A framework of the automatic grading system for mandarin proficiency is presented. The system included three modules which are pronunciation parameter database, evaluation model and automatic grading module. The PSO algorithm is utilized to adjust the connection weights of the selected ANN topology. Experimental results show the method performs better than the exiting methods.