A Security Event Management Framework Using Wavelet and Data-Mining Technique

As the rapid increase of network security events, network security monitoring and management on network behavior become more and more focused in the fields of computer science. This paper develops a kind of network security management framework using collection, analysis and integration, event-correlation and scenario-analysis technique to process the raw data gathered from hybrid network. Data-mining and wavelet technology are introduced into this framework, which can rapidly identify the types of security events from integrated information by using frequent episodes, wavelet and case-based reasoning. By these means, it can help to analyze the status of network security and then adjust the strategy to secure the network. We have implemented a prototype system following this framework, which can rapidly react to typical security events with visual output of result