Reasoning about fluids via molecular collections

  • Authors:
  • John W. Collins;Kenneth D. Forbus

  • Affiliations:
  • Department of Computer Science, University of Illinois;Department of Computer Science, University of Illinois

  • Venue:
  • AAAI'87 Proceedings of the sixth National conference on Artificial intelligence - Volume 2
  • Year:
  • 1987

Quantified Score

Hi-index 0.02

Visualization

Abstract

Hayes has identified two distinct ontologies for reasoning about liquids. Most qualitative physics research has focused on applying and generalizing his contained-liquid ontology. This paper presents a technique for generating descriptions using the molecular collection (MC) ontology, a specialization of his alternate ontology which represents liquids in terms of little "pieces of stuff" traveling through a system. We claim that MC descriptions are parasitic on the Contained-Stuff ontology, and present rules for generating MC descriptions given a Qualitative Process theory model using contained stuffs. We illustrate these rules using several implemented examples and discuss how this representation can be used to draw complex conclusions.