Mobile-agent based distributed fuzzy associative classification rules generation for OLAM

The capabilities of distribute data base to store huge amount of data and providing scalability, integrity leads to many of real time databases are stored in distributed nature. In order to apply data mining in real time applications it is important to provide efficient distributed data mining techniques. As a result of the use of online analytical mining (OLAM) technology in new fields of knowledge and the merging of data from different sources, it has become necessary for OLAM models to support distributed data mining technology. Data from different sources are not always consistent with the format and some sources may not reliable. In this paper we proposed a frame work for mobile-agent-based distributed analytical mining which can perform analytical mining on distributed and heterogeneous database systems and can manage reliable and unreliable sources differently. More over this architecture capable of performing mining on global and local data bases separately. Based on this architecture a flexible and efficient mobile-agent-based fuzzy association classification rules generation algorithm which can mine and present the global and local associative classification rules at the same time.