A Generic and Extendible Multi-Agent Data Mining Framework

A generic and extendible Multi-Agent Data Mining (MADM) framework, EMADS (the Extendible Multi-Agent Data mining System) is described. The central feature of the framework is that it avoids the use of agreed meta-language formats by supporting a system of wrappers. The advantage offered is that the system is easily extendible, so that further data agents and mining agents can simply be added to the system. A demonstration EMADS framework is currently available. The paper includes details of the EMADS architecture and the wrapper principle incorporated into it. A full description and evaluation of the framework’s operation is provided by considering two MADM scenarios. 1 Motivation and Goals Multi-Agent Data Mining (MADM) seeks to harness the general advantages of Multi-Agent Systems (MAS) in the application domain of Data Mining (DM). MAS technology has much to offer DM, particularly in the context of various forms of distributed and cooperative DM. The main issues with MADM are the disparate nature of DM and the wide range of tasks encompassed. Any desired generic MADM framework therefore requires a sophisticated communication mechanism to support it. In the work described here we address the communication requirements of MADM by using a system of mediators and wrappers coupled with an Agent Communication Language (ACL) such as FIPA ACL [8]. We believe this can more readily address the issues concerned with the variety and range of contexts to which a generic MADM can be applicable. The use of wrappers also avoids the need for agreed meta-language formats. To investigate and evaluate the expected advantages of wrappers and mediators in the context of generic MADM, we have developed and implemented (in JADE) a multi-agent framework, EMADS (the Extendible Multi-Agent Data mining System). The primary goal of the EMADS framework is extendibility; we wish to provide a means for integrating new DM algorithms and data sources in our MADM framework. However, EMADS also seeks to address some of the issues of DM that would benefit from the use of a generic framework. EMADS provides: – Flexibility in assembling communities of autonomous service providers, including the incorporation of existing applications. – Minimisation of the effort required to create new agents, and to wrap existing applications. – Support for end users to express DM requests without having detailed knowledge of the individual agents. The paper’s organisation is as follows. A brief review of some related work on MADM is presented in Section 2. The conceptual framework, together with an overview of the wrapper principle, is presented in Section 3 and Section 4. The framework’s operation is illustrated in Section 5 using two DM scenarios, and finally some conclusions are presented in Section 6.