Nowdays, with the development of the network technologies , the wide adoption of distributed computing environment , network security is very important. With the appearance of distributed Denial-of-Service(DDos)attacks etc.As a result distributed intrusion detection is an essential component of critical infrastructure protection method,but accumulated data are expoentinal increasing .traditional method can not complete this task.data mining .referred to as knowledge discovery in database, is the extraction of patterns representing valuable knowledge implicitly strored in large dataset. there are many techniques for datamining method. For example: Classification, Association, Clustering, Sequence.This pager analysis and compare , the distributed decision tree algorithms for classification and the distributed association rules algorithms. Advantages and disadvantages of these algorithms are also presented.I referenced these algorithms,also presented TFDM algorithms and TSPRINT algorithms in datamining of distributed intrusion detection.In the end, the effectiveness of two method is analysised and compared in an experiment, of course distributed intrusion detection source data is formatted by XML. Algorithms are developed by Microsoft Visual C++6. 0. From result data we can draw a conclusion.TFDM is better than TSPRINT in datamining of distributed large dataset.
|