Miscellaneous contributions to automated deductionJohn Slaney
Visualising Reasoning This is the abstract of the paper: John Slaney. Tools for graphical representation of problems in automated deduction or of proof searches are rare and mostly primitive. By contrast, there is a more substantial history of work in the constraint programming community on information visualisation techniques for helping programmers and end users to understand problems, searches and solutions. Here we consider the extent to which concepts and tools from a constraint programming platform can be adapted for use with automatic theorem provers.
MINLOG This is the abstract of the paper: John Slaney. Minlog is a theorem prover for propositional minimal logic and Heyting's intuitionist logic. It implements a decision procedure based on a cut-free sequent calculus formulation of these systems. While the method it uses is rather unsophisticated, on small problems MINLOG is fast. It achieves speed by being carefully coded (in C) and by eliminating many obvious redundancies in proof searches. It is thus useful as a point of comparison, since it represents what can be done by brute force rather than intelligence. The decision problem for the logics concerned is PSPACE-hard, so intelligence should easily triumph over mere speed. MINLOG provides a suitable baseline for evaluating implemented systems. NOTE:
Dr J K Slaney Phone (Aus.): (026) 125 8607 Theory Group Phone (Int.): +61 26 125 8607 Research School of Computer Science Fax (Aus.): (026) 125 8651 Australian National University Fax (Int.): +61 26 125 8651 Canberra, ACT, 0200, AUSTRALIA |