Production Systems as a Programming Language for Artificial Intelligence Applications. Volume I.
Abstract
This thesis develops a system architecture for artificial intelligence (AI), called production systems (PSs). Each production is a simple condition-action rule, with conditions stated on a global Working Memory and actions consisting primarily of simple modifications to that memory. Actions can also consist of forming new productions. PSs have been applied to a limited extent in computer science and to a somewhat larger extent to specialized studies in AI. Evaluations of PSs as a programming language are made according to the traits: practical feasibility, style, degree of theory-boundness, power and overhead of expression, productivity, efficiency, architectural flexibility, and level. PSs are found to be effective and advantageous for the programming constructs typical of AI systems.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1976
- Accession Number
- ADA037843
Entities
People
- Michael D. Rychener
Organizations
- Carnegie Mellon University