Mobile Agent- A Novel Paradigm for Data Mining

Abstract

The mobile agent paradigm has revolutionised the distributed computing environment. There are different approaches used in distributed computing. These are - client-server architecture, remote procedure architecture, code on demand architecture and mobile agent architecture. The client-server is based upon the concept of a server, that serves the various request of the clients and in remote procedure call approach, the machine can connect to another machine and retrieve the information remotely. The mobile agent technology is built upon the advancement in computing and communication technology. Further, it suits homogeneous as well as heterogeneous networks either using a wired or wireless media. To define it in simple words, we can say that - the mobile agent is an autonomous program that can migrates under its own or host control in the networks. Thus, it is simple, light in weight and performs the desired task efficiently, moreover this paradigm does not load the network and also the memory requirement is low. There are different programming languages used for writing programs that incorporate the mobile agent paradigm. Each of these languages has certain characteristics, which make them more suitable for a certain type applications that make use of mobile agent paradigm. Keeping in view the advantages and limitations of mobile agents it is beneficial for us to use the mobile agent paradigm. Data Mining is a process in which data can be extracted from different data warehouse based upon some pattern. Data Mining allows users to analyze data from many different perspective or angles, categorize it, and summarize the relationships. PMADE (Platform for Mobile Agent Distribution and Execution) is framework for large-scale multi-agent systems. The work presented makes use of mobile agent paradigm on PMADE platform for the process of data mining. The mobile agent paradigm is implemented on this platform and the process of data mining is performed and the desired output is presented.