Co-operative mobile and static agents for distributed communication systems for data mining applications

A mobile agent is a program that can travel during execution from one system to another system in a massive network. Mobile agent interacts with static agents and other resources to perform its task autonomously. Mobile agents are particularly attractive in distributed information retrieval applications. This paper discusses about how to parallelize the incremental algorithm for mining distributed dynamic datasets. This minimizes communication overhead between distributed systems and the central system. In this distributed approach, we use static agents at each distributed local system which are capable of generating local model (each static agent has a summary of its own database) as well as the global model (carries the summary of the entire data base) of the frequent item sets. This ability permits our system to generate high contrast frequent item sets, which allow us to examine how the data is positioned over different systems. Using the capabilities of the mobile agent, the knowledge could be retrieved from the local systems to the central system for making the communication process better and easier between the distributed systems.