Research On Data Mining Algorithms Of Information Security Classified Protection Evaluation

Information security classified protection evaluation launches a test or judgment on information systems, it is an important part of information security classified protection process. With the information security classified protection carrying out, evaluation task has become increasingly onerous, and how to improve the evaluation process is the problem to be solved.This paper focuses on how to improve the efficiency of the data mining work on evaluation data according to the data characteristics, which combines the data mining technology and the evaluation process better. Firstly, the feasibility of the data mining on evaluation data were analyzed and presented two models: cluster analysis and association analysis. In the association analysis aspect, this paper proposed a binary tire-based association rules mining algorithm BFPM by taking advantage of the features of datasets. This data structure makes the same item node lays in the same layer of the tree so that there is no need to create a frequent item head table. Compared with classic Fp-Growth algorithm, BFPM has the advantage of less occupied memory and a less complicated frequent pattern generate process.At last, the paper build an experimental platform based on tools like IBM data generator and JConsole,and conducted a series of experiments of BFPM argorithm with the exsisting Apriori argorithm and Fp-Growth argorithm. The experiments verified the binary tire based frequent pattern algorithm can meet the characteristics of the evaluation data, having a higher time and space efficiency.