| Under the guidance of the modern educational teaching concept which aims at cultivating innovative talents,the teaching mode of "student-centered" has become the focus of educators.In this model,the role of the teacher is changed to guide the students to carry out learning,and timely find and solve the problems students encounter in learning.To grasp the students’ knowledge structure accurately is the premise for teachers to play the above role.The knowledge structure of students includes group knowledge structure and individual knowledge structure.Students’ knowledge structure can be obtained through cognitive diagnosis.At present,most of the commonly used cognitive diagnostic models can accurately diagnose the individual knowledge structure,but cannot obtain the accurate group knowledge structure.However,in actual teaching,the accurate comprehensive group of students and individual cognitive diagnosis knowledge state,can make the teacher found the students cognitive defects in time,develop a targeted remedial teaching,guide students to learn better,at the same time,also can guide students individualized remedy according to the results of the individual diagnosis.By studying a large number of relevant literatures at home and abroad,the author found that the self-organizing clustering function of SOM neural network can be used to realize the cognitive diagnosis of students’ knowledge state.Therefore,this paper introduces SOM neural network into cognitive diagnosis and proposes a cognitive diagnosis model based on SOM neural network.Its basic idea is: through SOM self-organizing competitive neural network training is to respond to the actual pattern clustering,generate reaction model,then USES the expected response model for response pattern class,generated model class to the expected reaction model mapping,and then generate reaction model classes to ideal properties of nonlinear mapping,to determine the reaction model and the knowledge structure of the individuals in the class,the realization of ultimate state of knowledge of students cognitive diagnosis,student groups and individual knowledge structure.The cognitive diagnosis method based on SOM neural network can not only diagnose the individual knowledge state of students and provide support for personalized learning,but also accurately assess the cognitive difficulties of students,help teachers find the common problems among students and guide teachers to carry out targeted remedial teaching.This study will be as "introduction to database system" course content,the cognitive diagnosis model based on SOM neural network is applied to practical teaching to test its validity and feasibility of from diagnostic effect,real-time,and students’ learning effects and students’ satisfaction on the application of the model teaching four dimensions for data collection and statistical analysis.Through the analysis of relevant data,the following conclusions are drawn: the cognitive diagnosis model based on SOM neural network can accurately diagnose the knowledge structure of students and individuals in classroom teaching,and provide effective support for teachers to carry out targeted remedial teaching and students to conduct personalized learning.The model has good practicability,and teachers and students are satisfied with the application of cognitive diagnosis based on SOM neural network in practical teaching. |
