Accession Number : ADA300173

Title :   Steps Towards Terrain Knowledge Acquisition.

Descriptive Note : Final technical rept. Jul 93-Jul 94,

Corporate Author : ARGIALAS (DEMETRE P) ATHENS (GREECE)

Personal Author(s) : Argialas, Demetre P.

PDF Url : ADA300173

Report Date : 30 SEP 1995

Pagination or Media Count : 54

Abstract : In all previous approaches of the author and others in constructing prototype expert systems for landform interpretation, (1) knowledge related to the physiography of a site was expressed indirectly through a priori conditional probabilities of the occurrence of each landform in each physiographic section, (2) only a limited number of pattern elements were used, (3) there was no use of geomorphic process rules, (4) there was no use of regional context, (5) the systems processed only one landform at a time, and (6) there was no backward reasoning capability available. In this effort it has been developed a new conceptual scheme which was formalized through objects and rules so that to permit representation of (1) physiographic region reasoning, (2) an expanded (more detailed) set of-pattern elements, (3) geomorphic process reasoning, (4) regional context (spatial) reasoning, (5) multiple landform instances at a given interpretive scenario, (6) both forward and backward reasoning. The presented case studies concern typical terrain of the Basin and Range Province of Southwest USA. Detailed, 'took-level' knowledge pertaining to landforms and physiographic regions was gathered in the form of tables. The knowledge-base contains object and class representations of the typical landforms of the Basin and Range province such as alluvial fans, valley fills and playas as well as the typical physiographic regions of the province.

Descriptors :   *TERRAIN, *DATA ACQUISITION, *EXPERT SYSTEMS, *TERRAIN INTELLIGENCE, SCENARIOS, REASONING, PROBABILITY, PROTOTYPES, ALLUVIUM, REGIONS, CASE STUDIES, ARTIFICIAL INTELLIGENCE, GEOMORPHOLOGY, BASINS(GEOGRAPHIC), VALLEYS, LANDFORMS, TERRAIN MODELS, PLAYAS.

Subject Categories : Geology, Geochemistry and Mineralogy
      Computer Programming and Software

Distribution Statement : APPROVED FOR PUBLIC RELEASE