Modal logics extend classical logic with the modalities ”it is necessarily true that” and ”it is possibly true that” represented by the unary operators 2 and 3, respectively. First-order modal logics (FMLs) extend propositional modal logics by domains specifying sets of objects that are associated with each world, and the standard universal and existential quantifiers [7]. FMLs have many applications, e.g., in planning, natural language processing, program verification, querying knowledge bases, and modeling communication. These applications motivate the use of automated theorem proving (ATP) systems for FMLs. Whereas there are some ATP systems available for propositional modal logics, e.g., MSPASS [9] and modleanTAP [1], there were — until recently — no (correct) ATP systems that can deal with the full first-order fragment of modal logics. This abstract presents several new ATP systems for FML and sketches their calculi and working principles. The abstract also summarizes the results of a recent comparative evaluation of these new provers (see [4] for further details). The syntax of first-order modal logic adopted here is: F,G ::= P (t1, . . . , tn) | ¬F | F ∧ G | F ∨G | F ⇒ G | 2F | 3F | ∀xF | ∃xF . The symbols P are n-ary (n ≥ 0) relation constants which are applied to terms t1, . . . , tn. The ti (0 ≤ i ≤ n) are ordinary first-order terms and they may contain function symbols. The usual precedence rules for logical constants are assumed. Regarding semantics a well accepted and straightforward notion of Kripke style semantics for FML is adopted [7]. In particular, it is assumed that constants and terms are denoting and rigid, i.e. they always pick an object and this pick is the same object in all worlds. Regarding the universe of discourse constant domain, cumulative domain and varying domain semantics are considered. The following new ATP systems for FML were developed by the authors (partly as extensions of other systems); they support different combinations of modal logics and domain semantics (GQMLProver [19] has not been included since it returned incorrect results in our experiments for several formulae):
[1]
Jens Otten,et al.
The QMLTP Problem Library for First-Order Modal Logics
,
2012,
IJCAR.
[2]
Geoff Sutcliffe.
The TPTP Problem Library and Associated Infrastructure
,
2009,
Journal of Automated Reasoning.
[3]
Christoph Benzmüller,et al.
Implementing and Evaluating Provers for First-order Modal Logics
,
2012,
ECAI.
[4]
Lawrence C. Paulson,et al.
Quantified Multimodal Logics in Simple Type Theory
,
2009,
Logica Universalis.
[5]
Christoph Benzmüller,et al.
Combining and automating classical and non-classical logics in classical higher-order logics
,
2011,
Annals of Mathematics and Artificial Intelligence.
[6]
Lawrence J. Henschen,et al.
What Is Automated Theorem Proving?
,
1985,
J. Autom. Reason..
[7]
Geoff Sutcliffe,et al.
Automated Reasoning in Higher-Order Logic using the TPTP THF Infrastructure
,
2010,
J. Formaliz. Reason..
[8]
Torben Braüner,et al.
First-order modal logic
,
2007,
Handbook of Modal Logic.
[9]
Jens Otten.
Implementing Connection Calculi for First-order Modal Logics
,
2012,
IWIL@LPAR.
[10]
Lincoln A. Wallen,et al.
Automated deduction in nonclassical logics
,
1990
.
[11]
Bernhard Beckert,et al.
Free Variable Tableaux for Propositional Modal Logics
,
1997,
TABLEAUX.
[12]
Ullrich Hustadt,et al.
MSPASS: Modal Reasoning by Translation and First-Order Resolution
,
2000,
TABLEAUX.
[13]
G. Gentzen.
Untersuchungen über das logische Schließen. I
,
1935
.
[14]
Jens Otten,et al.
Clausal Connection-Based Theorem Proving in Intuitionistic First-Order Logic
,
2005,
TABLEAUX.
[15]
Christoph Kreitz,et al.
The ILTP Problem Library for Intuitionistic Logic
,
2007,
Journal of Automated Reasoning.
[16]
M. Fitting.
Proof Methods for Modal and Intuitionistic Logics
,
1983
.
[17]
Virginie Thion,et al.
A General Theorem Prover for Quantified Modal Logics
,
2002,
TABLEAUX.
[18]
Jens Otten.
Restricting backtracking in connection calculi
,
2010,
AI Commun..