The University of Edinburgh -
Division of Informatics
Forrest Hill & 80 South Bridge

MSc Thesis #9506

Title:Phoevos, Applying a Meta-Level Architecture to a Kbs for Harmonisation
Date: 1995
Abstract:"Which of its potential actions should an AI system perform at each point in the problem-solving process?" This is what is commonly known in the AI community as the control problem. The use of a meta-level architecture is a knowledge engineering approach that could tackle this problem. The key feature of this architecture is a declarative control language that allows one to specify the desired system behaviour by explicitly declaring the required control refine. This dissertation investigates the extent to which this suggestion is justified in practice. We first describe a language for explicitly specifying control strategies. This language forms part of a proposed domain-independent meta-level framework. Finally, starting from an existing system with a hard-wired control regime, a new system which was built as a case study for use of the architecture is described, PHOEVOS, a Knowledge Based System for harmonisation of music melodies.

[Search These Pages] [DAI Home Page] [Comment]