The Design Of A System For Designing Knowledge Representation Systems

The use of ebs t rec t dete types as a basis f o r designing experimental knowledge representa t ion systems is discussed. Abst rect dete types ere shown to heve features in common w i t h severe l d iverse representat ion formalisms ( e . g . semantic networks, frames and KLONE). For example, ebs t rec t data types have not ions analogous to concept, subconcept end i nhe r i t ance . The r e l a t i v e l y smal l conceptual d istence between abst rec t data types and knowledge representa t ion formalisms make them en i dea l veh ic le f o r implementing such formal isms.