A neural network-based systematic assessment of programming game interaction behavior of 0-6 year olds in an educational big data environment

Existing methods for assessing interaction behavior in programming games for young children have problems such as insufficient accuracy and lack of systematicity.How to scientifically assess the interaction behavior of young children in programming games and accurately grasp the characteristics of teacher-student verbal interaction has become the key to improving the quality of early childhood programming education.In this paper, we constructed an evaluation index system for the interaction behavior of programming games for 0-6 years old children, and proposed an evaluation model based on the Improved Particle Swarm Algorithm Optimized Radial Basis Neural Network (IPSO-RBF).The system contains two primary indicators, teacher speech behavior and student speech behavior, subdivided into 12 secondary indicators.Methodologically, nonlinear dynamic inertia weights and dynamic learning factors are used to improve the traditional PSO algorithm and optimize the width and weight parameters of the RBF neural network.The model performance is verified by 200 sets of training samples and 50 sets of test samples, and the results show that the test correct rate of IPSO-RBF neural network reaches 96%, and the average value of MSE is 0.121, which is significantly better than the 89.7% correct rate and 0.355 MSE value of PSO-RBF.Eleven programming teaching activities for 60 young children were analyzed for three types of instructional models: didactic, demonstrative, and socially constructive, and it was found that demonstrative classrooms had the highest number of significant sequences (27), followed by socially constructive (24), and didactic had the fewest ( 14).This study provides an effective tool for assessing the quality of early childhood programming education and is an important reference for optimizing teaching strategies.