A Logical Characterization of Forward and Backward Chaining in the Inverse Method
The inverse method is a generalization of resolution that can be applied to non-classical logics. We have recently shown how Andreoli’s focusing strategy can be adapted for the inverse method in linear logic. In this paper we introduce the notion of focusing bias for atoms and show that it gives rise to forward and backward chaining, generalizing both hyperresolution (forward) and SLD resolution (backward) on the Horn fragment. A key feature of our characterization is the structural, rather than purely operational, explanation for forward and backward chaining. A search procedure like the inverse method is thus able to perform both operations as appropriate, even simultaneously. We also present experimental results and an evaluation of the practical benefits of biased atoms for a number of examples from different problem domains.
KeywordsInverse Method Linear Logic Horn Clause Sequent Calculus Time Iters
Unable to display preview. Download preview PDF.
- 3.Cervesato, I., Pfenning, F., Walker, D., Watkins, K.: A concurrent logical framework I & II. Technical Report CMU-CS-02-101 and 102, Department of Computer Science, Carnegie Mellon University, 2002. Revised (May 2003)Google Scholar
- 7.Gentzen, G.: Untersuchungen über das logische Schließen. Mathematische Zeitschrift 39, 176–210 (1935): Szabo, M.E.: The Collected Papers of Gerhard Gentzen, pp. 68–131. North-Holland, Amsterdam (1969)Google Scholar
- 10.Jacob, M.: Howe. Proof Search Issues in Some Non-Classical Logics. PhD thesis, University of St. Andrews (September 1998)Google Scholar
- 16.Tammet, T.: Resolution, inverse method and the sequent calculus. In: Gottlob, G., Leitsch, A., Mundici, D. (eds.) KGC 1997. LNCS, vol. 1289, pp. 65–83. Springer, Heidelberg (1997)Google Scholar
- 17.Tamura, N.: Llprover. At: http://bach.istc.kobe-u.ac.jp/llprover