云环境下改进的非授权用户入侵行为分析及检测研究——以某高校学生专业倾向分析系统为例

Analysis and detection of intrusion behavior of non authorized users in cloud environment——Taking a college students' professional tendency analysis system as an example

为改进关联规则模块对非授权用户入侵行为分析的效率,并提高可靠性,提出以多尺度理论对关联规则挖掘进行辅助,根据概念分层理论来确定数据尺度和数据尺度划分,提出尺度下推的关联规则挖掘算法。该算法利用从大尺度数据集中得到的知识及多尺度数据集之间的关系,推导小尺度数据集中隐含的知识,而不对小尺度数据集进行直接挖掘,因此具有较高的运行效率。将该算法运用于云计算下基于改进的关联分析的非授权用户入侵行为分析模型,能有效提高检测速度。

In order to improve the efficiency of the intrusion behavior analysis of unauthorized users by associate rule modules, the paper indicates that the theory of multi-scale could assist association rule data mining, it presents the definition of data-scale-partition and data-scale based on the theory of concept hierarchy, and provides the scaling-push association rules mining algorithm. This algorithm uses the knowledge generated from large-scale data sets and the relationship between multi-scale data sets to deduce the embedded knowledge of small-scale data sets, rather than directly conduct data mining towards small-scale data sets, so it has high operating efficiency. Based on the improved association analysis, if the algorithm could be used in the cloud computing intrusion behavior analysis model of non authorized users, the detection speed should be increased effectively.