Qualitative engineering at various levels of conception design and evaluation of structures

  • Authors:
  • Bruno M. Franck

  • Affiliations:
  • Univ. of Minnesota, Minneapolis

  • Venue:
  • IEA/AIE '89 Proceedings of the 2nd international conference on Industrial and engineering applications of artificial intelligence and expert systems - Volume 1
  • Year:
  • 1989

Quantified Score

Hi-index 0.00

Visualization

Abstract

A multiple layer semantic net is proposed as a cognitive science framework to represent knowledge at various levels of abstractness. The multiple layers are qualitative at higher levels and quantitative at lower levels, and contain at all levels concepts that can be described by declarative and/or procedural statements. The semantic nets describe hierarchical knowledge associated with facts and events, and the procedures that are followed to process the information contained in the factual semantic nets. A Typicality Function is used to qualify the relevance of a concept to a class, and an Abstraction Function is used to describe at what level of representational abstraction the concept is described; both functions will be developed by combination of a fuzzy set-theory and knowledge-based prototype theory. All concepts and semantic net are represented by semiotic paradigms and manipulated as such. The information is processed through the multiple layer semantic net by qualitative reasoning.