Multiagent systems for large data clustering

Abstract

Multiagent system is an applied research area encompassing many disciplines. With increasing computing power and easy availability of storage devices vast volumes of data is available containing enormous amount of hidden information. Generating abstractions from such large data is a challenging data mining task. Efficient large data clustering schemes are important in dealing with such large data. In the current work we provide two different efficient approaches of multiagent based large pattern clustering that would generate abstraction with single database scan, integrating domain knowledge, multiagent systems, data mining and intelligence through agent-mining interaction. We illustrate the approaches based on implementation on practical data. © 2009 Springer-Verlag US.