Prediction of user behavioral intentions based on structural equation modelling in AI painting cognitive conditions

Abstract

The in-depth development of AI painting cognitive technology and the intelligent algorithm has made user intention analysis a hot spot in user research and also made the problem of willingness prediction the focus of psychology and economics. The original user intention method cannot solve the problem of user intention judgment under massive data, and the prediction accuracy of behaviour intention needs to be higher. Therefore, this paper proposes a structural variance model to analyze the user's behaviour willingness under the cognitive conditions of AI painting. Firstly, AI painting cognition is used to analyze user behaviour data, behaviour classification is carried out according to AI painting mental conditions, and irrelevant user intention data is deleted. Then, according to the user behaviour data classification results, it is compared with the previous user willingness analysis method. The different behaviour willingness domains are deeply excavated to output the behaviour willingness with a higher probability. MATLAB shows that the PLS-SEM structural equation model can improve the accuracy of user intention prediction; the accuracy rate reaches 86.5%, shorten the prediction time of user behaviour, and control it within 15 seconds to meet the needs of user behaviour prediction.