U. Hustadt and R. A. Schmidt. On evaluating decision procedures for modal logic. In M. E. Pollack, editor, Proc. of the Fifteenth Intern. Joint Conf. on Artificial Intelligence (IJCAI'97), pages 202--207. Morgan Kaufmann, 1997.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....generator for formulae in the modal logic K and tested their systems, Ksat and KsatC, on formulae produced by the generator. Unfortunately, some versions of their generator su#ered from problems which made the results of their testing suspect. The generator was modified by Hustadt and Schmidt[34, 35] to remove some of its problems and used to test the performance of their system, TA, a decision procedure for various propositional modal logics. We have created a new generator that extends the previous generators, and tested both Dlp and KsatC on the formulae produced by this new generator. ....

....being contradictory from just a single one of its clauses. When such formulae are used as input to a decision procedure that includes lexical normalisation, such as Dlp, many of the formulae are determined to be satisfiable or unsatisfiable just by lexical normalisation. Hustadt and Schmidt [34, 35] modified the generator of Giunchiglia and Sebastiani to eliminate some of these problems. The generator of Hustadt and Schmidt ensures that propositional variables only occur once in a 3CNFK (m) clause. In this way they eliminated many of the problematic formulae generated by Giunchiglia and ....

Hustadt, U. and R. A. Schmidt: 1997b, `On evaluating decision procedures for modal logic'. In: Proceedings of the 15th International Joint Conference on Artificial Intelligence (IJCAI-97), Vol. 1. pp. 202--207.

....generator for formulae in the modal logic K and tested their systems, Ksat and KsatC, on formulae produced by the generator. Unfortunately, some versions of their generator su#ered from problems which made the results of their testing suspect. The generator was modified by Hustadt and Schmidt[34, 35] to remove some of its problems and used to test the performance of their system, TA, a decision procedure for various propositional modal logics. We have created a new generator that extends the previous generators, and tested both Dlp and KsatC on the formulae produced by this new generator. ....

....being contradictory from just a single one of its clauses. When such formulae are used as input to a decision procedure that includes lexical normalisation, such as Dlp, many of the formulae are determined to be satisfiable or unsatisfiable just by lexical normalisation. Hustadt and Schmidt [34, 35] modified the generator of Giunchiglia and Sebastiani to eliminate some of these problems. The generator of Hustadt and Schmidt ensures that propositional variables only occur once in a 3CNFK (m) clause. In this way they eliminated many of the problematic formulae generated by Giunchiglia and ....

Hustadt, U. and R. A. Schmidt: 1997a, `On evaluating decision procedures for modal logic'. Technical Report MPI-I-97-2-003, Max-Planck-Institut Fur Informatik, Im Stadtwald, D 66123 Saarbrucken, Germany.

....checking there is large and rapidly expanding body of experimental knowledge; see, e.g. 7] In contrast, empirical aspects of modal satisfiability checking have only recently drawn the attention of researchers. We now have a number of test sets, some of which have been evaluated extensively [4, 10, 8, 13, 12]. In addition, we also have a clear set of guidelines for performing empirical testing in the setting of modal logic [10, 12] Currently, there are three main test methodologies for modal satisfiability solvers, one based on hand crafted formulas, the other two based on randomly generating ....

U. Hustadt and R. A. Schmidt. On Evaluating Decision Procedures for Modal Logic. In Proc. IJCAI-97, pages 202--207, 1997.

....1 class machine. FaCT also takes about 200 seconds for this task on a comparable machine. DLP has also been tested on two sets of benchmarks, the Tableaux 98 comparison benchmarks [ Heuerding and Schwendimann, 1996 ] and a collection of hard random modal formulae due to Hustadt and Schmidt [ Hustadt and Schmidt, 1997 ] DLP outperformed the existing description logic systems on the Tableaux 98 benchmarks and was competitive with FaCT, the fastest description logic system on the random formulae. More detail on DLP s performance is available in a paper discussing the performance of the various optimisations in ....

U. Hustadt and R. A. Schmidt. On evaluating decision procedures for modal logic. Technical Report MPI-I-97-2-003, MaxPlanck -Institut Fur Informatik, Im Stadtwald, D 66123 Saarbrucken, Germany, February 1997.

....logics, the better option would be to develop a variety of different algorithms and heuristics, each performing well on a limited class of inputs. For example, it is known that restricting the number of propositional atoms to be used or the depth of modal nesting may reduce the complexity (cf. [22, 24, 18, 34]) Also, when considering specific applications it is possible to reduce some of the infinitary character of collective beliefs and intentions to more manageable proportions (cf. 16, Ch. 11] In this paper we leave out temporal considerations. Our full theory is, however, based on Kripke models ....

Hustadt, U., Schmidt, R.: On Evaluating Decision Procedures for Modal Logics, Proceedings IJCAI'97 (M. Pollack, Ed.), Morgan Kauffman, Los Angeles (CA), 1997.

.... rule is an instance of a LSR restricted to literals and formulae in CNF; ffl DPLL is an instance of a sequent calculus with two forms of LSR (restricted to literals) for CNF; ffl The fi c i rules of KE [ D Agostino and Mondadori, 1994 ] modus ponens of HARP and the Gamma simp 0;1 rules in [ Hustadt and Schmidt, 1997 ] are instances of LSR applied only to the main connectives of tableaux formulae; ffl propositional hyper tableaux [ Baumgartner et al. 1996 ] are an instance of the sequent calculus with LSR and a selection function for CNF; ffl for modal logics basic KSAT [ Hustadt and Schmidt, 1997 ] is a ....

.... in [ Hustadt and Schmidt, 1997 ] are instances of LSR applied only to the main connectives of tableaux formulae; ffl propositional hyper tableaux [ Baumgartner et al. 1996 ] are an instance of the sequent calculus with LSR and a selection function for CNF; ffl for modal logics basic KSAT [ Hustadt and Schmidt, 1997 ] is a restricted sequent calculus for logic K with only propositional LSR for modal CNF. 3 Experimental Results For random 3 CNF a simple LSR produces already the easy hard easy pattern. Only in the transition region the analogous of DPLL split performs better. Beside CNF benchmarks we have ....

U. Hustadt and R. Schmidt. On evaluating decision procedure for modal logic. In Proc. of IJCAI-97, 1997.

....logics, the better option would be to develop a variety of different algorithms and heuristics, each performing well on a limited class of inputs. For example, it is known that restricting the number of propositional atoms to be used or the depth of modal nesting may reduce the complexity (cf. [15, 16,24]) Also, when considering specific applications it is possible to reduce some of the infinitary character of collective beliefs and intentions to more manageable proportions (cf. 13, Ch. tt] In this paper we leave out temporal considerations. Our full theory is, however, based on Kripke models ....

U. Hustadt and R. Schmidt. On evaluating decision procedures for modal logics. In M. Pollack, editor, Proceedings IJCAI'97, Los Angeles (CA), 1997. Morgan Kauffman.

....sequent calculus (such as tableaux) are hopeless for real life satisfiability and validity search. Even for decidable problems with a natural appeal for tableaux, such as modal or propositional logics, experimental results have shown that other algorithms outperform them by orders of magnitude [9, 18, 24, 34, 42]. Such a huge gap leads to think that this is not a matter of efficient implementations but rather of theoretical nature. The key question is whether such a gap is inherent to tableaux methods or something is simply missing. This answer is not only of theoretical relevance but has an extreme ....

....be shaded by many black box steps. To fill the computational gap, a number of works have improved the effectiveness of tableaux method by switching from ground calculi to free variables versions [31, 16] with smart skolemisation techniques [2] adding ad hoc rules for modus ponens and tollens [10, 24, 28], imposing regularity conditions [10, 25] using controlled form of cut [25, 10] or factoring and merging [27, 41] exploiting universal variables [7, 4] incorporating features of hyper resolution [3] They have improved the efficiency of the calculus but sometimes at the expenses of the ....

[Article contains additional citation context not shown here]

U. Hustadt and R. A. Schmidt. On evaluating decision procedure for modal logic. In Proceedings of the 15th International Joint Conference on Artificial Intelligence (IJCAI-97), 1997. To appear.

....comparison scale (p search simulation) It shown that KEM p search simulates SST while SST cannot p search simulate KEM. 1 Introduction In the last few years several comparisons (competition) for theorem provers for modal logic have been held and experimental research has been carried out (cf. [11, 10]) Despite the potential interest for future possible applications, we believe, that they provided little or no insight on better theoretical architecture for modal theorem provers, since external factors could influence the performance. In this paper two labelled approaches to modal tableaux are ....

Ullrich Hustadt and Renate Schmidt. On evaluating decision procedures for modal logic. In IJCAI'97, volume 1, pages 202--207, 1997.

....been against test suites for propositional modal logics, using the propositional modal logic interface for DLP. We have tested against the test suite for the Tableaux 98 propositional modal logic comparison [8] and against a collection of random formulae initially generated by Hustadt and Schmidt [10] . The Tableaux 98 test suite consists of several classes of formulae (e.g. branch) in both provable and non provable forms, for each of K, KT, and S4. For each class of formula, 21 examples of supposedly exponentially increasing difficulty are automatically generated from a basic pattern which ....

....the oldestfirst heuristic enhances backjumping. Our second propositional modal logic test suite uses a common method for testing SAT decision procedures [4] that has been adapted for use with propositional modal K by Giunchiglia and Sebastiani [7] and further refined by Hustadt and Schmidt [10] . The method uses a random generator to produce formulae, with the characteristics of the formulae being controlled by a number of parameters. Each formula is a conjunction of L K clauses, where a K clause is a disjunction of K elements, each element being negated with a probability of 0.5. An ....

U. Hustadt and R. A. Schmidt. On evaluating decision procedures for modal logic. Technical Report MPI-I-97-2-003, Max-Planck-Institut Fur Informatik, Saarbrucken, Germany, February 1997.

No context found.

U. Hustadt and R. A. Schmidt. On evaluating decision procedures for modal logic. In M. E. Pollack, editor, Proc. of the Fifteenth Intern. Joint Conf. on Artificial Intelligence (IJCAI'97), pages 202--207. Morgan Kaufmann, 1997.

No context found.

U. Hustadt and R. A. Schmidt. On evaluating decision procedures for modal logic. Research report MPI-I-97-2-003, Max-Planck-Institut fur Informatik, Saarbrucken, Germany, February 1997.

No context found.

U. Hustadt and R. A. Schmidt. On evaluating decision procedures for modal logics. In M. E. Pollack, editor, Proceedings of the Fifteenth International Joint Conference on Artificial Intelligence (IJCAI '97), pages 202--207, Morgan Kaufmann, 1997.

No context found.

Hustadt, U., Schmidt, R.: On Evaluating Decision Procedures for Modal Logics, Proceedings IJCAI'97 (M. Pollack, Ed.), Morgan Kauffman, Los Angeles (CA), 1997.

No context found.

U. Hustadt and R. A. Schmidt. On evaluating decision procedures for modal logic. In Proc. IJCAI-97, pages 202--207, 1997.

No context found.

U. Hustadt and R. Schmidt. On evaluating decision procedures for modal logic. In Proceedings IJCAI-97, pages 202--207, 1997.

No context found.

U. Hustadt and R. A. Schmidt. On evaluating decision procedures for modal logic. In Proceedings IJCAI-97, pages 202--207, 1997.

No context found.

Hustadt, U. and R. Schmidt, On evaluating decision procedures for modal logic, in: Proc. of the 15th Iternational Joint Conference on Artificial Intelligence (IJCAI'97), 1997, pp. 202--207.

No context found.

U. Hustadt and R.A. Schmidt. On Evaluating Decision Procedures for Modal Logic. In Proc. IJCAI-97, pages 202--207, 1997.

No context found.

U. Hustadt and R.A. Schmidt. On Evaluating Decision Procedures for Modal Logic. In Proc. IJCAI-97, pages 202--207, 1997.

No context found.

U. Hustadt and R. A. Schmidt. On evaluating decision procedures for modal logic. In Proceedings of the 15th International Joint Conference on Artificial Intelligence (IJCAI-97), volume 1, pages 202--207, 1997.

No context found.

U. Hustadt and R. A. Schmidt. On evaluating decision procedures for modal logic. In Proceedings of the 15th International Joint Conference on Artificial Intelligence (IJCAI-97), Volume 1, pp. 202--207, 1997.

No context found.

Hustadt, U., Schmidt, R.: On Evaluating Decision Procedures for Modal Logics, Proceedings IJCAI'97 (M. Pollack, Ed.), Morgan Kauffman, Los Angeles (CA), 1997.

No context found.

U. Hustadt and R. A. Schmidt. On evaluating decision procedures for modal logic. In Proceedings of the 15th International Joint Conference on Artificial Intelligence (IJCAI-97), volume 1, pages 202--207, 1997.

No context found.

Hustadt, U. and R. A. Schmidt, On Evaluating Decision Procedures for Modal Logic, in: Proc. IJCAI-97, 1997, pp. 202--207.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC