Research On The Application Of Data Mining Technology In The Evaluation And Decision Of Network Security Level Protection

Network security level protection is an important work to protect information and information carriers according to the importance level.In the network security level protection workflow,the network security level protection evaluation is an important part of it.In recent years,with the continuous advancement of my country’s network security grade protection work,the following problems have appeared in the network security grade protection evaluation process: First of all,due to the characteristics of the evaluation work,it is necessary for the evaluation personnel to enter the evaluation site to conduct on-site evaluation work.In this process,due to subjective or objective reasons,the assessors may misjudge the assessment results;Secondly,the grade protection evaluation work will be carried out regularly,and a large amount of evaluation data will be continuously generated.However,there is no valuable analysis of these historical data,and the existing data has not been effectively used to form a reference guide for the subsequent grade protection evaluation work.The research focus of this article is to use data mining technology to analyze the ten security aspects of graded protection evaluation,solve the problems that appear in the current grade protection evaluation,and obtain substantive conclusions that are helpful to the evaluation work,and provide better evaluation personnel.Support decision-making guidance.The main research results of this paper are as follows:First of all,for the problem of misjudgment of evaluation results,this paper studies and analyzes the classic algorithm of association rules,and conducts in-depth research on the Apriori algorithm and the FP-growth algorithm.After theoretical analysis,the performance of the two algorithms is known.Combined with the characteristics of the evaluation data and experimental verification,the performance of the FP-growth algorithm is better than that of the Apriori algorithm.Through the analysis of association rules,the relationship between the ten safety-level non-conformance evaluation items in the evaluation index system is obtained.Through the analysis of the association relationship between the non-conformance evaluation items and the actual evaluation status,the rationality of the association rules is proved.The finally obtained association rules can assist the assessors in the assessment work,reducing the misjudgment of the assessment results.Secondly,for the problem of forming a reference for evaluation,this paper adopts the clustering analysis algorithm.First,it studies and analyzes the basic ideas of the classic clustering algorithm K-means algorithm and hierarchical clustering AGNES algorithm,and proposes the use of the sum of squared errors(Sum Of The Squared Errors,SSE)to preliminarily determine the value of k in the K-means algorithm,and because the determination of the elbow point of SSE is not rigorous,it is proposed to calculate the slope between two points to determine the final value of k.Through the analysis of the algorithm and the characteristics of the evaluation data,because the evaluation data has fewer outliers,it is not obvious that the hierarchical clustering algorithm is good at processing outlier data.On the contrary,it will be caused by the excessive amount of calculation.The algorithm runs slowly,so we finally consider using the K-means algorithm for mining and analysis for clustering analysis.Finally,by clustering the scores of ten security levels,four clustering results are obtained,and corresponding security strategies are proposed according to the characteristics of each category.