Research on Bayesian Based Network Precision Computation Model

Due to the booming development of e-commerce information, the number of website users has significantly increased. How to use user access traces on e-commerce platforms to conduct big data analysis of customer behavior and achieve precise product promotion has attracted much research interest in e-commerce. However, in the current research methods and technical means, there are many technical challenges, especially in the effectiveness of data analysis, information, image modeling, and data processing. Thus, it is necessary to analyze website customer attributes, behavioral characteristics, and browsing records using big data technology. In this study, Bayesian network models were used to explore customer behavioral characteristics and estimate the posterior probability of potential product generation or consumption behavior and usage behavior. The Bayesian network-based recommendation algorithm proposed in this study showed an accuracy of about 20% higher than traditional product recommendation models.