International Journal of Scientific and Research Publications, Volume 4, Issue 6, June 2014 1 ISSN 2250-3153 www.ijsrp.org An Analysis on Multi-Agent Based Distributed Data Mining System R.HEMAMALINI *, Dr.L.JOSEPHINE MARY ** * Research scholar St.Peter’s University ** Professor & HOD, MCA Dept. Sri Ram Engineering College,Veppampet Abstract- The Distributed Data Mining (DDM) is a branch of the field of data mining that offers a framework to mine distributed data paying careful attention to the distributed data and computing resources. Usually, data-mining systems are designed to work on a single dataset. On the other hand with the growth of networks, data is increasingly dispersed over many machines in many different geographical locations. Also, even as most practical data-mining algorithms operate over propositional representations are known as first order learning. In existing system, the concept of knowledge is very important in data mining. In order to get the correct knowledge from the data mining system, the user must define the objective and specify the algorithms and its parameters exactly with minimum effort. If the data mining system produces large number of meaningful information by using a specialized data mining algorithm like association, clustering, decision trees etc., it will take more time for the end- users to choose the appropriate knowledge for the problem discussed. Even choosing the correct data mining algorithm involves more time for the system. Developing a data mining system that uses specialized agents with the ability to communicate with multiple information sources, as well as with other agents requires a great deal of flexibility. The main objective of this paper titled on “An Analysis on Multi-Agent Based Distributed Data Mining System“ describes the knowledge integration, Knowledge Integration in Distributed Data-Mining and Heterogeneous vs. Homogeneous Data-Mining, a literature survey of Multi-Agent Based Distributed Data Mining System, a Model Of Multi –Agent System Based Data Mining, the improving DDM performance by combining distributed data mining and multi-agent system and Data Mining using Multiple Agents. Index Terms- MAS – Multi agent System, DDM- Distributed Data Mining, DMA – Data Mining Agent , ILP- Identify Local Pattern, KI – Knowledge Integration. I. INTRODUCTION Data-mining or Knowledge Discovery is concerned with extracting knowledge from databases and/or knowledge bases using machine learning techniques. The first order learning is to enables us to explore the aspects of knowledge integration and theory refinement which do not appear in propositional systems. Software has the response to the problem of using the vast amounts of information stored on networked systems. There are many types of software agent; however, agents are typically thought of as being 'intelligent' programs which have some degree of self-sufficiency. We intend to design an open, flexible data-mining agent. A group of these agents will be able to co- operate to discover knowledge from distributed sources. In DDM, one of the two assumptions is commonly adopted as to how data is distributed across sites: homogeneously (horizontally partitioned) and heterogeneously (vertically partitioned). Both viewpoints adopt the conceptual idea that the data tables at each site are partitions of a single global table. In the homogeneous case, the global table is horizontally partitioned . The tables at each site are subsets of the global table; they have exactly the same attributes. In the heterogeneous case the table is vertically partitioned, each site contains a collection of columns as sites do not have the same attributes. However, each tuple at each site is assumed to contain a unique identifier to facilitate matching. It is important to stress that the global table viewpoint is strictly conceptual. It is not necessarily assumed that such a table was physically realized and partitioned to form the tables at each site. The development of data mining algorithms that work well under the constraints imposed by distributed datasets has received significant attention from the data mining community in recent years. Local computation is done on Improving DDM Performance by Combining Distributed Data Mining and Multi- Agent System. The Multi-Agent Learning is a number of co-operative distributed learning systems have been produced. Each agent has a data-source and a clustering algorithm. The agents propose rules which characterize the data seen and review other agents' proposals. Eventually consent about the knowledge extracted from the data is reached. Each agent has local knowledge and either an inductive or deductive learning algorithm. Agents attempt to solve a problem-solving task by either retrieving the knowledge required, or by using learning to acquire it. Failures result in communication with other agents which are passed sub- goals, which are then treated as tasks. There are three ways learning can occur when data is distributed. These relate to when agents communicate with respect to the learning process: • The first approach gathers the data in one place. The distributed database management systems is used to provide a single set of data to an algorithm is an example of this. The problem with such an approach is that it does not make efficient use of the resources usually associated with distributed computer networks.