一种物联网安全属性概念提取方法  

Concept Retrieval Method for Security Attributes of Internet of Things

作　　者：马政朝 郑瑞娟[1] 吴庆涛[1] 张明川[1] 

机构地区：[1]河南科技大学电子信息工程学院,河南洛阳471023

出　　处：《计算机仿真》2014年第3期303-307,共5页Computer Simulation

基　　金：国家自然科学基金(61142002;U1204614;61003035);河南省科技厅资助项目(.112102210187)

摘　　要：研究物联网安全属性的风险精确评估问题,传统的划分方法不能有效反映数据间的实际分布规律和边界划分,使得最终不能得到令人容易理解的关联识别。针对物联网安全数据分布的随机性,运用云变换从定量数据中获取数据的范围和分布规律,将精确的数值转换成恰当的定性描述,在此基础上提出了一种梯形云概念提升算法。实验结果表明,运用云变换算法可以有效地将数量型属性的定义域划分为多个云的定性概念,对划分结果进行合理提升,最终得到易理解的、有效的关联规则,输出的概念层次更符合实际,为物联网安全的准确评估提供了依据。The security attributes of Internet of things （IOT） quantitative association rules method is the premise for accurately assessing IOT risk. But traditional classification methods cannot effectively reflect the actual distribution of data or have orderliness boundary, finally can not get easily understandable related knowledge. In this paper, based on the randomness of IOT security data distribution, the cloud transformation was used to extract qualitative concept from quantitative data and divert the precise numerical value into appropriate qualitative description. On the basis of this, a new algorithm of trapezium cloud concept was proposed. Experimental results show that using cloud transformation algorithm can effectively divide quantitative attribute domain into multiple qualitative concept based cloud, the classification results are more reasonable. Finally, understandable and effective association rules were ob- tained easily, which makes the output concept hierarchy closer to reality.

关 键 词：物联网 概念提取 云变换 梯形云 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]