CAKE – Classifying, Associating and Knowledge DiscovEry - An Approach for Distributed Data Mining (DDM) Using PArallel Data Mining Agents (PADMAs)

This paper accentuate an approach of implementing distributed data mining (DDM) using multi-agent system (MAS) technology, and proposes a data mining technique of ldquoCAKErdquo (classifying, associating & knowledge discovery). The architecture is based on centralized parallel data mining agents (PADMAs). Data mining is part of a word, which has been recently introduced known as BI or business intelligence. The need is to derive knowledge out of the abstract data. The process is difficult, complex, time consuming and resource starving. These highlighted problems addressed in the proposed model. The model architecture is distributed, uses knowledge-driven mining technique and flexible enough to work on any data warehouse, which will help to overcome these problems. Good knowledge of data, meta-data and business domain is required for defining rules for data mining. Taking into consideration that the data and data warehouse has already gone through the necessary processes and ready for data mining.