Data Security Risk Assessment and Control based on Improved Apriori and NSGA-II

Abstract

With the rapid development of information technology, the prominence of net work data security risks is escalating. Accurately evaluating these risks and devising effective control strategies have emerged as urgent imperatives. To address this challenge, this study introduces an enhanced version of the Apriori algorithm to discern correlations within data and perform corresponding risk assessments. Additionally, it employs the elitist non-dominated sorting genetic algorithm for data security risk control. The findings demonstrate that the probability of occurrence of primary risk factors, as calculated using the enhanced Apriori, exceeds 0.7, with approximately 75% confidence in the association between each factor and medium to high risk. These calculations align closely with the ”main network security risks” outlined in the unit’s 2021 network security work summary. Furthermore, the distribution of Pareto optimal solution sets derived from the multi-objective evolutionary algorithm based on decomposition exhibits non-uniformity, whereas those obtained from the elitist non-dominated sorting genetic algorithm manifest diverse and evenly distributed outcomes. Moreover, considering risk state values and control costs, this study adopts a third approach to mitigate security risks concerning application systems and data. This strategy yields a risk level of only 0.386 and incurs a cost consumption of merely 2,404,800 yuan. The proposed data security risk assessment and control strategy demonstrate strong feasibility, effectively enhancing the value of data utilization and delivering practical benefits to enterprises or organizations.