海量加密军用数据下的频繁项目集挖掘仿真

Huge Amounts of Encrypted Military Simulation Data of Frequent Itemsets Mining

军用加密数据为了达到保密的目的,人为设定了较多的加密规则,打破了数据之间常规的关联性。在进行军用数据挖掘建立关联规则时,由于数据关联规则被人为隐藏,递归生成关联条件模式树的过程中,传统的FP-tree算法挖掘算法,由于加密数据的关联复杂性,会递归生成大量条件模式树,导致后期挖掘过程占用了大量的挖掘算法资源,挖掘效率较低。提出基于改进FP-tree的海量加密军用数据下频繁项目集挖掘算法,依据海量加密军用数据下频繁项目集挖掘原理,在FP-tree算法的基础上,依据预剪枝策略减少挖掘节点,通过单向有序FP-tree防止每次存储当前挖掘出的频繁项目集之前都需要超集检验,建立项目表格,避免递归生成条件模式树浪费资源。将提出的改进FP-tree算法应用到海量加密军用数据下频繁项目集的挖掘中,获取的实验结果说明,改进FP-tree算法在提高加密军用数据频繁项目集挖掘速度及准确率方面具有较高的优越性。

Military encrypted data in order to achieve the purpose of confidentiality, artificial more encryption rules, broke the conventional correlation between data. In the military to establish data mining association rules, as the data association rules by artificial hidden, recursive generated tree model in the process of the associated condi- tions, the traditional mining algorithms FP - tree algorithm, due to the associated complexity of encrypted data, re- cursive model tree generates a large number of conditions, lead to late mining algorithm for mining process takes up a lot of resources, the mining efficiency is low. Is proposed based on improved FP - tree under the massive encrypted military data of frequent itemsets mining algorithms, based on mass data encrypted military under the principle of fre- quent itemsets mining, on the basis of FP - tree algorithm, on the basis of preliminary node pruning strategy to re- duce mining, orderly through one-way FP - tree to prevent each store before the current mining the frequent itemsets need superset test, establishing the project forms, avoid recursive generation condition model tree waste of resourees. Will improve FP - tree algorithm is applied to the massive encrypted military data under frequent itemsets mining, obtained the result of the experiment shows that improve FP - tree algorithm to improve encrypted military data fre- quent itemsets mining has the advantages of high speed and accuracy.