Knowledge structures in UC, the UNIX Consultant

  • Authors:
  • David N. Chin

  • Affiliations:
  • University of California, Berkeley, Berkeley, CA

  • Venue:
  • ACL '83 Proceedings of the 21st annual meeting on Association for Computational Linguistics
  • Year:
  • 1983

Quantified Score

Hi-index 0.02

Visualization

Abstract

The knowledge structures implemented in UC, the UNIX Consultant are sufficient for UC to reply to a large range of user queries in the domain of the UNIX operating system. This paper describes how these knowledge structures are used in the natural language tasks of parsing, inference, planning, goal detection, and generation, and how they are organized to enable efficient access even with the large database of an expert system. The structuring of knowledge to provide direct answers to common queries and the high usability and efficiency of knowledge structures allow UC to hold an interactive conversation with a user.