Using our work on Terrain Map Understanding (Section 17.3), we plan to build the expert system support in MovieScript to be used in late vision tasks such as proximity analysis, GIS knowledge-based processing, and object recognition. This project, also a part of the ASAS Map separates program, is planned with Coherent Research, Inc., Syracuse NY, where a similar expert system capability is being developed for analyzing black-and-white handmade maps used by the local electric company (Niagara Mohawk).
We are also planning to build the knowledge-based ``intelligent'' CASE tools to enforce economy and to accelerate the MOVIE development process. Typical examples include smart-class browsers or automated interface builders based on ``fuzzy'' specification of user requests. This approach is in the spirit of the Knowledge Based Software Engineering (KBSE) technology, recently advocated by DARPA on the basis of comprehensive analysis of software costs [Boehm:90a] as the efficient economy measure for the next generation software processes. Implementation of the KBSE concepts requires integrating expert system techniques with conventional software engineering practices. Since PostScript derives from Lisp, its appropriate extension in MovieScript towards symbolic processing offers a natural integration platform for KBSE tools.