Machine learning models increasingly influence decisions in high-stakes settings such as finance, law and hiring, driving the need for transparent, interpretable outcomes. However, while explainable approaches can help understand the decisions being made, they may inadvertently reveal the underlying proprietary algorithm—an undesirable outcome for many practitioners. Consequently, it is crucial to balance meaningful transparency with a form of recourse that clarifies why a decision was made and offers actionable steps following which a favorable outcome can be obtained. Counterfactual explanations offer a powerful mechanism to address this need by showing how specific input changes lead to a more favorable prediction. We propose Model-Agnostic Causally Constrained Counterfactual Generation (MC3G), a novel framework that tackles limitations in the existing counterfactual methods. First, MC3G is model-agnostic: it approximates any black-box model using an explainable rule-based surrogate model. Second, this surrogate is used to generate counterfactuals that produce a favourable outcome for the original underlying black box model. Third, MC3G refines cost computation by excluding the “effort” associated with feature changes that occur automatically due to causal dependencies. By focusing only on user-initiated changes, MC3G provides a more realistic and fair representation of the effort needed to achieve a favourable outcome. We show that MC3G delivers more interpretable and actionable counterfactual recommendations compared to existing techniques all while having a lower cost. Our findings highlight MC3G’s potential to enhance transparency, accountability, and practical utility in decision-making processes that incorporate machine-learning approaches.

Abstract Assurance cases offer a structured way to present arguments and evidence for certification of systems where safety and security are critical. However, creating and evaluating these assurance cases can be complex and challenging, even for systems of moderate complexity. Therefore, there is a growing need to develop new automation methods for these tasks. While most existing assurance case tools focus on automating structural aspects, they lack the ability to fully assess the semantic coherence and correctness of the assurance arguments. In prior work, we introduced the Assurance 2.0 framework that prioritizes the reasoning process, evidence utilization, and explicit delineation of counter-claims (defeaters) and counter-evidence. In this paper, we present our approach to enhancing Assurance 2.0 with semantic rule-based analysis capabilities using common-sense reasoning and answer set programming solvers, specifically s(CASP). By employing these analysis techniques, we examine the unique semantic aspects of assurance cases, such as logical consistency, adequacy, indefeasibility, etc. The application of these analyses provides both system developers and evaluators with increased confidence about the assurance case.

Abstract This paper proposes a new methodology for early validation of high-level requirements on cyber-physical systems with the aim of improving their quality and, thus, lowering chances of specification errors propagating into later stages of development where it is much more expensive to fix them. The paper presents a transformation of a real-world requirements specification of a medical device—the Patient-Controlled Analgesia (PCA) Pump—into an Event Calculus model that is then evaluated using Answer Set Programming and the s(CASP) system. The evaluation under s(CASP) allowed deductive as well as abductive reasoning about the specified functionality of the PCA pump on the conceptual level with minimal implementation or design dependent influences and led to fully automatically detected nuanced violations of critical safety properties. Further, the paper discusses scalability and non-termination challenges that had to be faced in the evaluation and techniques proposed to (partially) solve them. Finally, ideas for improving s(CASP) to overcome its evaluation limitations that still persist as well as to increase its expressiveness are presented.

Abstract The development of large language models (LLMs), such as GPT, has enabled the construction of several socialbots, like ChatGPT, that are receiving a lot of attention for their ability to simulate a human conversation. However, the conversation is not guided by a goal and is hard to control. In addition, because LLMs rely more on pattern recognition than deductive reasoning, they can give confusing answers and have difficulty integrating multiple topics into a cohesive response. These limitations often lead the LLM to deviate from the main topic to keep the conversation interesting. We propose AutoCompanion, a socialbot that uses an LLM model to translate natural language into predicates (and vice versa) and employs commonsense reasoning based on answer set programming (ASP) to hold a social conversation with a human. In particular, we rely on s(CASP), a goal-directed implementation of ASP as the backend. This paper presents the framework design and how an LLM is used to parse user messages and generate a response from the s(CASP) engine output. To validate our proposal, we describe (real) conversations in which the chatbot’s goal is to keep the user entertained by talking about movies and books, and s(CASP) ensures (i) correctness of answers, (ii) coherence (and precision) during the conversation—which it dynamically regulates to achieve its specific purpose—and (iii) no deviation from the main topic.

Abstract Automated commonsense reasoning (CR) is essential for building human-like AI systems featuring, for example, explainable AI. Event calculus (EC) is a family of formalisms that model CR with a sound, logical basis. Previous attempts to mechanize reasoning using EC faced difficulties in the treatment of the continuous change in dense domains (e.g. time and other physical quantities), constraints among variables, default negation, and the uniform application of different inference methods, among others. We propose the use of s(CASP), a query-driven, top-down execution model for Predicate Answer Set Programming with Constraints, to model and reason using EC. We show how EC scenarios can be naturally and directly encoded in s(CASP) and how it enables deductive and abductive reasoning tasks in domains featuring constraints involving both dense time and dense fluents.

Abstract Building Information Modeling (BIM) produces three-dimensional object-oriented models of buildings combining the geometrical information with a wide range of properties about materials, products, safety, to name just a few. BIM is slowly but inevitably revolutionizing the architecture, engineering, and construction industry. Buildings need to be compliant with regulations about stability, safety, and environmental impact. Manual compliance checking is tedious and error-prone, and amending flaws discovered only at construction time causes huge additional costs and delays. Several tools can check BIM models for conformance with rules/guidelines. For example, Singapore’s CORENET e-Submission System checks fire safety. But since the current BIM exchange format only contains basic information about building objects, a separate, ad-hoc model pre-processing is required to determine, for example, evacuation routes. Moreover, they face difficulties in adapting existing built-in rules and/or adding new ones (to cater for building regulations, that can vary not only among countries but also among parts of the same city), if at all possible. We propose the use of logic-based executable formalisms (CLP and Constraint ASP) to couple BIM models with advanced knowledge representation and reasoning capabilities. Previous experience shows that such formalisms can be used to uniformly capture and reason with knowledge (including ambiguity) in a large variety of domains. Additionally, incorporating checking within design tools makes it possible to ensure that models are rule-compliant at every step. This also prevents erroneous designs from having to be (partially) redone, which is also costly and burdensome. To validate our proposal, we implemented a preliminary reasoner under CLP(Q/R) and ASP with constraints and evaluated it with several BIM models.

Understanding the meaning of a text is a fundamental challenge of natural language understanding (NLU) research. An ideal NLU system should process a language in a way that is not exclusive to a single task or a dataset. Keeping this in mind, we have introduced a novel knowledge driven semantic representation approach for English text. By leveraging the VerbNet lexicon, we are able to map syntax tree of the text to its commonsense meaning represented using basic knowledge primitives. The general purpose knowledge represented from our approach can be used to build any reasoning based NLU system that can also provide justification. We applied this approach to construct two NLU applications that we present here: SQuARE (Semantic-based Question Answering and Reasoning Engine) and StaCACK (Stateful Conversational Agent using Commonsense Knowledge). Both these systems work by ``truly understanding'' the natural language text they process and both provide natural language explanations for their responses while maintaining high accuracy.

Abstract Logic programming with tabling and constraints (TCLP, tabled constraint logic programming) has been shown to be more expressive and in some cases more efficient than LP, CLP, or LP + tabling. Previous designs of TCLP systems did not fully use entailment to determine call/answer subsumption and did not provide a simple and well-documented interface to facilitate the integration of constraint solvers in existing tabling systems. We study the role of projection and entailment in the termination, soundness, and completeness of TCLP systems and present the design and an experimental evaluation of Mod TCLP, a framework that eases the integration of additional constraint solvers. Mod TCLP views constraint solvers as clients of the tabling system, which is generic w.r.t. the solver and only requires a clear interface from the latter. We validate our design by integrating four constraint solvers: a previously existing constraint solver for difference constraints, written in C; the standard versions of Holzbaur’s and , written in Prolog; and a new constraint solver for equations over finite lattices. We evaluate the performance of our framework in several benchmarks using the aforementioned solvers. Mod TCLP is developed in Ciao Prolog, a robust, mature, next-generation Prolog system.

Abstract Extending ASP with constraints (CASP) enhances its expressiveness and performance. This extension is not straightforward as the grounding phase, present in most ASP systems, removes variables and the links among them, and also causes a combinatorial explosion in the size of the program. Several methods to overcome this issue have been devised: restricting the constraint domains (e.g., discrete instead of dense), or the type (or number) of models that can be returned. In this paper we propose to incorporate constraints into s(ASP), a goal-directed, top-down execution model which implements ASP while retaining logical variables both during execution and in the answer sets. The resulting model, s(CASP), can constrain variables that, as in CLP, are kept during the execution and in the answer sets. s(CASP) inherits and generalizes the execution model of s(ASP) and is parametric w.r.t. the constraint solver. We describe this novel execution model and show through several examples the enhanced expressiveness of s(CASP) w.r.t. ASP, CLP, and other CASP systems. We also report improved performance w.r.t. other very mature, highly optimized ASP systems in some benchmarks.

Logic programming systems with tabling and constraints (TCLP, tabled constraint logic programming) have been shown to be more expressive and in some cases more efficient than those featuring only either tabling or constraints. Previous implementations of TCLP systems which use entailment to determine call / answer subsumption did not provide a simple, uniform, and well-documented interface to facilitate the integration of additional constraint solvers in existing tabling systems, which would increase the application range of TCLP. We present the design and an experimental evaluation of Mod TCLP, a framework which eases this integration. Mod TCLP views the constraints solver as a client of the tabling system. The tabling system is generic w.r.t. the constraint solver and only requires a clear, small interface from the latter. We validate our design by integrating four constraint solvers: a re-engineered version of a previously existing constraint solver for difference constraints, written in C; the standard versions of Holzbauer's CLP(Q) and CLP(R), written in Prolog; and a new constraint solver for equations over finite lattices. We evaluate the performance of our framework in several benchmarks using the aforementioned constraint solvers. All the development work and evaluation was done in Ciao Prolog.