Model generation with Boolean Constraints
Miyuki Koshimura, Hiroshi Fujita, Ryuzo Hasegawa · Institutional Repositories DataBase (IRDB) · 2001
We present a simple method for eliminating redundant searches in model generation. The method employs Boolean Constraints which are conjunctions of ground instances of clauses having participated in proofs. Boolean Constraints work as sets of lemmas with which duplicate subproofs and irrelevant model extensions can be eliminated. The method has been tentatively implemented on a constraint logic programming system. We evaluated effects of the method by proving some typical problems taken from the TPTP problem library.