Personalized teaching of college psychology based on information fusion technology

Abstract

Abstract This paper explores the cognitive attention mechanism in mental activities, analyzes the learning emotion contained therein as well as the information of thinking activity that carries out the whole cognitive process of mental learning, and obtains the information construction data of the three-dimensional interest model after the estimation of the learner’s head gesture and the recognition of dynamic expression. A conditional random forest model is built using multimodal information fusion technology, and a natural smile detection method is proposed based on it. After training, it realizes the information estimation of the learner’s head pose data, generates the smiley face classifier based on conditional random forest, and determines the teaching decision boundary using K-Means clustering. To analyze the psychological personalized teaching results of the model, an empirical study is conducted through a controlled experiment. The experimental results show that the learning efficiency of the model on the CelebA dataset and SMILEsmileD dataset is improved, the accuracy rate is stabilized at 95% after the number of iterations 10, and the model’s performance is superior. The majority of the students in the experiment have a mastery of psychological knowledge that is around 0.85, and there is a significant positive correlation, and personalized teaching has a more significant effect.