A Knowledge-Based Multi-Agent System for Geospatial Data Conflation

In this paper, we apply software agent technology paradigm to a well-known problem of Geographical Information Systems (GIS) known as conflation. Conflation is the complex process of recognizing and removing inconsistencies in geographical feature datasets that are stored in multiple databases. Specifically, we show how the technological advantages of an agent system can be combined with expert system techniques to provide a feasible system architec ture for distributed conflation. Our work utilizes expert systems in the multi-agent based infrastructure to perform a distributed conflation algorithm. Here, we present our system design, which utilizes both static and mobile agents.