Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.652392
Title: Using middle-out reasoning to guide inductive theorem proving
Author: Hesketh, Jane Thurmann
Awarding Body: University of Edinburgh
Current Institution: University of Edinburgh
Date of Award: 1992
Availability of Full Text:
Access through EThOS:
Full text unavailable from EThOS. Please try the link below.
Access through Institution:
Abstract:
Techniques derived from proof theory for logic alone have been insufficient as a basis for efficient, elegant automatic theorem proving. They concentrate on syntax, neglecting both strategy for particular domains and classes of problem, and guidance from modelling human mathematicians. A novel technique suggested by Bundy, developing ideas from Ernst & Newell, is to reason 'middle-out'. Often, the overall structure of a proof may be known, but its details must be fleshed out according to the individual theorem. Conventional research might use heuristic guidance to backtrack over all possibilities. Middle-out reasoning uses variables as place-holders for parameters still to be chosen. These place-holders become instantiated by the requirements of the subsequent proof. Decisions which would multiply the search space are postponed until more information is available. This is an exciting development in search control, making extensive use of strategic guidance and harnessing tools from human reasoning. This thesis reports research on its use for the synthesis of tail recursive functions from corresponding naive functions and for proofs requiring generalisation. It enables the development of a unified framework for generalisation. An existing proof planning and development system based on Martin-Lö Type Theory, Oyster/CLAM, is used as vehicle, augmented by higher order unification.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.652392  DOI: Not available
Share: