Data Modeling and Data Warehousing Techniques to Improve Intrusion Detection

摘要: This chapter describes data mining and warehousing techniques that can improve the performance usability of Intrusion Detection Systems (IDS). Current IDS do not provide support for historical analysis summarization. presents to model network traffic alerts using a multi-dimensional star schemas. was used perform security detect denial service attacks. Our also be handle heterogeneous sources (e.g. firewall logs, system calls, net-flow data) enable up two orders magnitude faster query response times analysts as compared current state art. We have our implement prototype is being successfully at Army Research Labs. has helped analyst in detecting intrusions generating reports on trend analysis.