Study On The Application Of Data Mining Based On Intrusion Detection System

With the application and popularization of computer technology and wide use of network technology, network security issues become increasingly prominent, as they pose a huge security risk to the network itself and network-based information systems, and force people to study, research, and gradually solve such issues. Network intrusion detection technology is another one of the key network security technologies after such traditional protection means as firewall, data encryption, etc., it can provide essential security protection by identifying malicious behavior against computer itself and network resources. In this technology, data mining has, as a machine learning application in database, been successfully applied in the network intrusion detection and will give full play to the advantage of its ability to handle large amounts of data and improve the efficiency and accuracy of intrusion detection.Of course, there are still such problems in most current intrusion detection systems as lack of effectiveness, adaptability, scalability, higher omission rate or false alarm rate, too much reliance on the human experts, and taking up much system resources.Connecting the status quo of network security and beginning with the concept of invasion, this paper gradually researches and analyzes intrusion detection, intrusion detection systems, network intrusion detection systems, data mining, data mining approach, data mining in intrusion detection and other related theories and technologies. Based on this, this paper makes a brief analysis and preliminary study for FADT-based network intrusion detection system model by learning from experts, scholars and research institutions on the use of data mining methods and by adopting academic achievements of quickly building the decision tree (FADT) method in network intrusion detection. Finally, this paper introduces research direction and goals of this systematic study, combining with the current development trend of intrusion detection systems.The intrusion detection system has, based on FADT (autonomous rapid decision trees), fully drawn on the benefits of other intrusion detection systems, and thus can not only cope, in the network intrusion detection, with massive data analysis in the current high-speed broadband networks with fairly low the omission rate and false alarm rate, but also has adaptivity and scalability. All this enables it to make real-time response against large-scale network attacks, and find hidden intrusion in time in the normal use.