Research Of Snort Intrusion Detection System Based On BMHS4C And M-Apriori

With the diversification of cybersecurity threat and the pervasive of hackertechnology, more and more intrusion methods can easily cross the traditional firewall tosteal important personal information and business secrets. Intrusion detection system is agood solution to these problems, but it still have some defects such as missing report、false report and processing delay.In order to find relevant information hidded from a variety of data packet, this paperuse data mining technology. Data mining algorithms can search out the correlationinformation hided from database by the item set. Through combining the data mining andintrusion detection, we can not only eliminate a lot of useless data, but also produce newrules to provide the lasting vitality of the intrusion detection, improve the accuracy of theintrusion detection.In this paper,the data mining decision tree algorithm C4.5and improved Apriorialgorithm are applied to the intrusion detection system, and the pattern matchingalgorithm of intrusion detection system is also improved. By decreasing the amount ofdata and updating the rule base itself, improve the efficiency of pattern matching. At lastthrough experiments prove the decrease of false alarm rate and missing report rate, andthe improvement of detection efficiency in the improved intrusion detection system.