概念格在入侵检测中的应用研究

Application research of concept lattice in intrusion detection

为了发现潜在的、有效的入侵检测规则,提高入侵检测系统的检测率,将概念格与入侵检测技术相结合提出了一种基于概念格理论的入侵检测系统。系统通过对收集的数据进行预处理、数据规范化,使用属性约简得到最小属性集构建概念格,同时分析概念间的蕴涵关系,获得非冗余的分类规则。基于概念格的入侵检测模型与其它检测方法相比要求的训练数据获取简单,实验结果表明,使用该模型减少了实现分类的运算量,提高了入侵检测的检测率,有效控制了检测的误检率。

In order to discover potential and effective intrusion detection rules, intrusion detection system based on concept lattice is given. The system includes data preprocessing, data standardization, attribute reduction, productionofconceptlattice. The implication relations between concepts are analyzed, the non-redundant classification rules are extrated. The intrusion detection model based on concept lattice compared with other detection methods easy access to training data. The experiment shows that the use of a model reduces the amount of computing, improves the accuracy of intrusion detection, effectively controls the false rate of detection.