Schubert's steamroller problem: formulations and solutions
Journal of Automated Reasoning
Journal of Automated Reasoning
A-SATCHMORE: SATCHMORE with availability checking
New Generation Computing
Positive Unit Hyperresolution Tableaux and Their Application to Minimal Model Generation
Journal of Automated Reasoning
SATCHMO: A Theorem Prover Implemented in Prolog
Proceedings of the 9th International Conference on Automated Deduction
Non-Horn Magic Sets to Incorporate Top-down Inference into Bottom-up Theorem Proving
CADE-14 Proceedings of the 14th International Conference on Automated Deduction
R-UNSEARCHMO: A Refinement on UNSEARCHMO
PRICAI '02 Proceedings of the 7th Pacific Rim International Conference on Artificial Intelligence: Trends in Artificial Intelligence
SATCHMOREBID: SATCHMO(RE) with BIDirectional relevancy
New Generation Computing
Eliminating redundant search space on backtracking for forward chaining theorem proving
Journal of Computer Science and Technology
An Improvement of Herbrand's theorem and its application to model generation theorem proving
Journal of Computer Science and Technology
UNSEARCHMO: eliminating redundant search space on backtracking for forward chaining theorem proving
IJCAI'01 Proceedings of the 17th international joint conference on Artificial intelligence - Volume 1
Hi-index | 0.01 |
We introduce a method for reducing the redundant search space for SATCHMO's model generation approach by means of intelligent backtracking. During the reasoning, we mark an asserted consequent atom as “useful” whenever it has been used as an antecedent atom for forward chaining. We show that a splitting of the consequence of a non-Horn clause is unnecessary if one of its consequent atoms is found not to be “useful” at the time it is retracted from the database on backtracking, and therefore the remaining splitting over the clause's consequence can be immediately abandoned. In this way, much of the redundant search space can be eliminated. Our method is simple in principle, easy to implement in Prolog, independent of other refinements, and effective for model generation theorem proving.