Knowledge representation on the connection machine

  • Authors:

  • M. Evett;L. Spector;J. Hendler

  • Affiliations:

  • Dept. of Computer Science, University of Maryland, College Park, MD;Dept. of Computer Science, University of Maryland, College Park, MD;Dept. of Computer Science, University of Maryland, College Park, MD

  • Venue:
  • Proceedings of the 1989 ACM/IEEE conference on Supercomputing
  • Year:
  • 1989

Quantified Score

Hi-index 0.00

Visualization

Abstract

A primary motivation for the development of the Connection Machine (CM) was to create a vehicle for artificial intelligence research. The original design was largely based upon Fahlman's NETL machine [Fah79], the primary purpose of which was to effect large semantic networks, a paradigm of artificial intelligence. To date, however, only a small amount of AI research is being conducted on the CM. Discounting neural net and computer vision research, the amount is miniscule. The lack of AI tools for the CM is a primary cause for this dearth of research. AI researchers proposing to use the CM must first develop the necessary AI tools before beginning work on their projects. For example, there are no existing inference systems, knowledge representation packages, expert system toolkits, etc. PARKA, a pseudo-acronym for “Parallel Knowledge Representation and Association”, was developed as the prototype of one such tool: a knowledge representation system modeled on a frame-based representation language (FDL) [Tou87] paradigm.