A distributed intelligent geographical modeling environment for spatial decision support

Often in geography, researchers and decision-makers have to deal with semi- or unstructured spatial problems. The evaluation of sites for ecological restoration is one example of such a problem. To help answer questions of this nature, Spatial Decision Support Systems (SDSS) were developed. SDSS integrates geographically referenced data and spatial models to generate and analyze alternative solutions that can be evaluated by decision-makers. In order to support the decision-maker in the process of decision-making, the SDSS must be easy-to-use and flexible.;Current SDSS, however, are either not user-friendly, or support only a limited number of decision-making environments. Significant programming effort, coupled with GIS software expertise, is required to create and modify an SDSS. Further, when a GIS expert creates or modifies an SDSS, it becomes a "black box" whose inner workings are often unknown to the decision-maker.;The Distributed Intelligent Geographical Modeling Environment (DIGME) framework was created to overcome this limitation, and assist decision-makers integrate data and models into an SDSS. DIGME uses an artificial intelligence concept called "agents" to mimic the behavior of a GIS expert. The agents achieve this task using three key technologies and an innovative method to integrate models and data.;The three key technologies are: (i) Metadata---information about available data and models; (ii) Blackboard architecture---a mechanism for reasoning about data transformations; and (iii) Geographic Data Definition Language---a language to specify search parameters for spatial data.;The innovative methods employed by the agents to integrate data and models involve (i) searching the Internet for spatial models and geographic data using metadata; (ii) converting available data into the format required by the models (Blackboard architecture is used to determine necessary data transformations); and (iii) running models for the decision-maker and presenting the results.;A prototype SDSS, designed to be both user-friendly and flexible, was developed using the DIGME framework to evaluate ecological and economic impacts of agricultural policy for the Cache River watershed of southern Illinois.