Application of data mining in intrusion detection

This paper analyzing the lastest research progress and mian problems existed of IDS, researching advantage of data mining technique applied to IDS, and analyzing disadvantage of IDS based on data mining technique, for the problem of time and space inefficient in intrusion detection based on data mining, and aims at the research of frequential pattern algorithm, inproved frequential pattern algorithm, used two-step growth model instead of one step model to speed up the mining speed, and increase in time features, attribute related and axis attribute to constraint, and the experimental results show the proved algorithm improve the time and space efficiency, decrease the time of scan database and generate less meaningless pattern, heighten the availability for rule.