The Role of Agents in Distributed Data Mining: Issues and Benefits

  • Authors:
  • Matthias Klusch;Stefano Lodi;Gianluca Moro

  • Affiliations:
  • -;-;-

  • Venue:
  • IAT '03 Proceedings of the IEEE/WIC International Conference on Intelligent Agent Technology
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

The increasing demand to extend data mining technology to data setsinherently distributed among a large number of autonomous andheterogeneous sources over a network with limited bandwidth hasmotivated the development of several approaches to distributed datamining and knowledge discovery, of which only a few make use ofagents. We briefly review existing approaches and argue for thepotential added value of using agent technology in the domain ofknowledge discovery, discussing both issues and benefits. We alsopropose an approach to distributed data clustering, outline itsagent-oriented implementation, and examine potential privacyviolating attacks in which agents may incur.