Abstract
Employing the expert system shell discussed in part 1 of this two–part article, three simple geographical expert systems are constructed as didactic examples. The first deals with land–type classification with remotely–sensed data, the second with climatic classifications with regular data files, and the third involves discussions on building DTM–related expert systems. All are rule–based expert systems easily built from the shell. It is apparent that the shell provides a powerful software environment for developing knowledge–based GIS. Directions for further research are also outlined.

This publication has 5 references indexed in Scilit: