Exception Date Mining And Applied Research In Web Service Security

Important topics in Data Mining are the discovery of interesting and useful knowledge. Which is implied, potentional and unknown previously. While exception knowledge is extremely special case of data pool and is distinguished from other data. Exception knowledge reveals objects' abnormal rule deviated from the common and describes the differentiation. So exception knowledge is also named deviation knowledge. Server Logs may consist of some special Outlier data, which be -longs to exception knowledge.Considering of Web service security problem, we made use of the Data Mining technology united serve logs to explore and analysis the Outlier, based on the statistics method and the distance method. Furthermore, from the Outlier we probe some malicious business and proposed the concrete Data Mining method. According to testing, it has been proved that this method is available to examine the special Outlier data and malicious business in the Web Service.