To solve the problem that most computer network security assessment systems cannot comprehensively analyze data, this article proposes a research method for computer network confidentiality information security system based on the Bayesian classification algorithm. This research takes the computer network security secret-related information security as the core, cleverly uses the EBCA back-propagation model, and comprehensively uses nonlinear functions to propose a computer network secret-related information security application system considering EBCA, which effectively solves the complex cross-linking relationship. The effectiveness of the method is verified by simulation experiments. Simulation experiments show that when the network performance is in the optimal state, the slope is 1, the intercept is 0, and the fitting degree is 1. The method proposed in this article is especially suitable for the nonlinear characteristics of computer network security and can accurately, comprehensively, and systematically reflect the security operation status of network security and confidential information. Then, through the simulation test, it is found that the method in this study can effectively meet the error requirements of equipment integrity and equipment normal rate.