期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2016
卷号：83
期号：2
出版社：Journal of Theoretical and Applied
摘要：Due to the exponential rise of the network attacks and increasing development of software tools and techniques for intrusion detection, the rule based intrusion detection system has become an essential solution for real-time anomaly detection. Basically, traditional data mining based intrusion detection methods generate a large set of predefined patterns most of them are high false rate and inaccurate. There is a need to optimize the real-time network attacks due to the variation in new attack type, instance set and attributes. To address the issue of high false rate and dynamic data integration, a new anomaly detection system using data mining model has been proposed to find the real time DOS/DDOS patterns by integrating network packets capturing from different systems on the network and kdd99 dataset. This system generates intrusion patterns by integrating the predefined attacks and new attacks as early as possible with low false rate. Experimental results show that proposed dynamic model optimizes the real-time true positive patterns with high accuracy compared to traditional models.
