answer set programming

Existing Answer Set Programming (ASP) systems struggle to flexibly support rapid prototyping of diverse non-monotonic temporal logics, such as Temporal Equilibrium Logic (TEL), Modal Epistemic Logic (MEL), and Dynamic Epistemic Logic (DEL). This work proposes a unified declarative metaprogramming framework that extends the clingo theory language with a type system and nested modal guards to ensure semantic correctness during grounding and enable uniform encodings of these logics. By integrating a metaprogram transformation pipeline, stable model semantics preservation, Fischer–Ladner closure handling, and interval constraint management, the framework yields a general-purpose tool called metasp. This tool substantially enhances the extensibility and development efficiency of temporal ASP systems, offering a cohesive infrastructure for experimenting with and implementing a broad class of non-monotonic temporal reasoning formalisms.

Industrial problems in dynamic domains—such as scheduling, path planning, and production sequencing—impose both temporal and metric constraints, yet existing Answer Set Programming (ASP) frameworks lack native support for reasoning over time and quantitative measures. Method: This paper proposes the first systematic extension of ASP integrating dynamic, temporal, and metric logics into both its syntax and semantics. We design an extended ASP syntax supporting timestamps and finite-domain clocks, introduce temporal rule encoding mechanisms, and employ incremental solving strategies to jointly enable declarative modeling and efficient time-sensitive inference. Contribution/Results: Our framework establishes a theoretically grounded and computationally tractable foundation for dynamic ASP. Experiments on standard dynamic benchmarks demonstrate a 40% improvement in inference efficiency and millisecond-scale response times. A deployable, industrial-grade prototype for dynamic decision-making has been implemented, bridging expressive modeling with real-time computational feasibility.

To address three key challenges in Answer Set Programming (ASP) formal verification—lack of modularity, grounding dependence, and tool limitations—this paper introduces a novel semantic framework grounded in Here-and-There (HT) logic and multi-sorted first-order logic (FOL). Our method establishes, for the first time, a verifiable semantic mapping from ASP rules to multi-sorted FOL, enabling rule-level independence and data-agnostic semantics without requiring grounding. This framework supports modular modeling and compositional verification, and is compatible with mainstream automated theorem provers (e.g., Vampire, SPASS). Experimental evaluation demonstrates end-to-end automatic verification of semantic properties for full classes of ASP rules; on standard benchmarks, it achieves significant improvements in scalability and composability. The approach thus establishes a new paradigm for trustworthy verification of AI systems grounded in declarative logic programming.

First-order Answer Set Programming (ASP) lacks systematic support for modular and parameterized subroutines, hindering structured program design. This work proposes a formal framework for parameterized modular logic programs, introducing for the first time a parameterization mechanism and intensional declarations into modular ASP. By means of precise semantic mappings, it faithfully captures the collective control mechanisms employed in clingo. The approach establishes a theoretical foundation for modular ASP while enabling declarative definition, reuse, and clear semantics for modules. Consequently, it effectively bridges the gap between modularity and traditional ASP, demonstrating strong expressiveness and practical utility in program structuring and instantiation.

This work addresses the challenge of automatically translating natural language into Answer Set Programming (ASP) as a critical step toward robust neuro-symbolic systems. It introduces ASP-Bench, the first multidimensional reasoning benchmark encompassing core ASP features—such as choice rules, aggregates, and optimization statements—with 128 carefully curated natural language–ASP pairs. The authors propose a feedback-driven iterative modeling paradigm grounded in the ReAct framework, which leverages solver feedback in a closed loop to progressively refine generated logic programs. Evaluated on ASP-Bench, this approach achieves full saturation, demonstrating both the efficacy of the proposed paradigm and enabling fine-grained analysis of the factors governing translation difficulty.

This work addresses the lack of effective static analysis techniques for goal-directed Answer Set Programming (ASP), which hinders compile-time verification and optimization. It introduces abstract interpretation to this paradigm for the first time, proposing a top-down analysis algorithm based on PLAI fixpoints and designing a novel Shared-Constraints abstract domain to precisely capture variable relationships induced by constraints. Implemented as a preprocessor in Ciao Prolog and integrated into the s(CASP) system, the approach enables efficient detection of spurious odd loops, optimized forall evaluation, and global constraint simplification. These capabilities substantially enhance both the compile-time analyzability and runtime performance of goal-directed ASP programs.

This work investigates the computational complexity and efficient solving techniques for second-order Quantified Answer Set Programming with weak constraints (2-ASP(Q)ʷ). It provides the first complete characterization of the complexity of its primary reasoning tasks within the Δ₃^P level of the polynomial hierarchy. The paper introduces a Counterexample-Guided Abstraction Refinement (CEGAR) solving strategy specifically tailored to the semantics of ASP(Q), effectively addressing non-trivial cases not handled by prior approaches. Implemented in the Casper system, the proposed method demonstrates significant performance improvements over existing solvers on multiple challenging benchmark instances, thereby confirming its effectiveness and practical applicability.

This work addresses the inefficiency of solving combinatorial problems in Answer Set Programming (ASP) by introducing, for the first time, a large language model (LLM)-based streamlining approach adapted from constraint programming. By designing targeted prompts, the method guides the LLM to automatically generate semantically diverse and non-redundant structured constraints. These candidate constraints are dynamically filtered through ASP encoding analysis and validity verification, enabling the construction of virtually optimal encoding variants across multiple benchmarks. Evaluated on three ASP competition benchmarks, the approach achieves up to a 4–5× speedup in solving time, significantly enhancing computational efficiency and demonstrating the effectiveness and generalizability of LLM-driven streamlining in ASP.

This work addresses the challenges novice learners face when acquiring Answer Set Programming (ASP), which stem from the language’s high level of abstraction and excessive structural freedom. To lower the entry barrier, we present EZASP—the first Visual Studio Code extension integrating the Easy ASP methodology. EZASP guides users toward writing well-structured and safe ASP programs through structured code snippets, automatic program reordering, and a configurable real-time feedback mechanism. The system implements core features including syntax error highlighting, unsafe variable detection, and structural validation. Empirical user studies demonstrate that EZASP significantly reduces the learning curve for beginners and enhances both programming efficiency and code correctness for both novice and experienced users.

This work addresses the limited expressiveness of current Constraint Answer Set Programming (CASP) solvers when handling numerical constraints, which often forfeits key features of Answer Set Programming (ASP), such as default reasoning, undefined attributes, non-deterministic choices, and aggregates. To bridge this gap, the paper introduces FLINGO, a novel language that, for the first time, integrates ASP-style declarative modeling of numerical attributes into linear integer constraints. FLINGO achieves this by employing a syntactic translation mechanism that compiles its constructs into standard CLINGCON input. The approach preserves the efficiency of state-of-the-art numerical constraint solving while substantially enhancing modeling expressiveness. A prototype implementation demonstrates the effectiveness and practicality of the proposed method across several illustrative examples.