Deductive Search for Errors in Free Data Type Specifications using Model Generation
Paper in proceeding, 2002

The presented approach aims at identifying false conjectures about free data types. Given a specification and a conjecture, the method performs a search for a model of an according counter specification. The model search is tailor-made for the semantical setting of free data types, where the fixed domain allows to describe models just in terms of interpretations. For sake of interpretation construction, a theory specific calculus is provided. The concrete rules are ‘executed’ by a procedure known as model generation. As most free data types have infinite domains, the ability of automatically solving the non-consequence problem is necessarily limited. That problem is addressed by limiting the instantiation of the axioms. This approximation leads to a restricted notion of model correctness, which is discussed. At the same time, it enables model completeness for free data types, unlike approaches based on limiting the domain size.

Author

Wolfgang Ahrendt

Chalmers, Department of Computing Science

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

03029743 (ISSN) 16113349 (eISSN)

Vol. 2392 211-225
978-3-540-43931-8 (ISBN)

Subject Categories

Computer and Information Science

DOI

10.1007/3-540-45620-1_18

ISBN

978-3-540-43931-8

More information

Created

10/8/2017