Research On The Integrated Association Rule Mining System

this paper adopt a new method to integrate existing association rule mining methods, and propose an integrated association rule data mining system with a style of its own. The system include three aspects: generalized association rule mining, generalized sequential pattern mining, remining on the result of the above mentioned. This paper discusses these three aspects in detail. Content includes: ⑴ The association mining system include three aspect above mentioned. Generalized association rule mining shows an association relationship of a time section in database; generalized sequential pattern shows data association relationship with time proceeding. remining works in a three-dimensional space composed of time subject and action. ⑵This paper mainly research on a approach of generalized association rule mining. This method can break away form the limited of mining from single dataset. In this way we can acquire more useful information. ⑶Generalized sequential mining have comparability with generalized association rule mining and can learn much from the later. The difference between them is that sequential mining conform to time-attribution. According to the experience of the practical, while things record in the database with the same time label, there also have some order relationship hidden in the records. This paper take advantage of these messages and make sequential mining perfect. ⑷Remining is a new filed of data mining. It includes four parts: First one is to use association mining method behind association mining. It uses a new method to find the relationship between the patterns that we get in the first step. The second one is use association method after sequential mining. The meaning of the second one is to find the trending of the rules which we get from sequential mining. The third one use association rule mining method behind sequential mining. This method can used to identify the difference between different object-sequence. The last one use sequential mining method behind sequential mining. This provide us a new way to run sequential mining itself.