A Security Assessment Strategy for Corporate Financial Systems Based on Data Mining Techniques

Abstract

Abstract The security assessment of corporate financial systems has become a popular area of research, although few results have been published so far. This research focuses on integrating the association rule algorithm with Bayesian networks to evaluate the security of corporate financial systems. It primarily employs Bayesian networks to calculate the conditional probability, prior joint probability, and posterior probability of each attribute node in the IDRI-Tree to assess the security of the company’s financial system. Additionally, abnormal data from Company A’s financial system were collected to evaluate the method’s effectiveness. The results indicate that Company A’s financial system has data discrepancy compliance rates of 98.92% and 96.50%, respectively. The security assessment scores for monitoring, evaluation, and assessment in financial system security are the lowest, averaging 1.41, while the scores for payment, service, and support are the highest, averaging 3.32. The financial system security assessment method proposed in this paper demonstrates high practical value and offers a reference for the future establishment and improvement of financial system security assessment methods.