Auflistung nach:
Auflistung Künstliche Intelligenz 38(1-2) - August 2024 nach Titel
1 - 10 von 13
Treffer pro Seite
Sortieroptionen
- ZeitschriftenartikelAn ASP Implementation of Defeasible Deontic Logic(KI - Künstliche Intelligenz: Vol. 38, No. 0, 2024) Governatori, GuidoWe present a novel implementation of Defeasible Deontic Logic as an Answer Set Programming meta-program, and we evaluate the performance of the implementation against a recent set of benchmarks.
- ZeitschriftenartikelChallenges for Non-Classical Reasoning in Contemporary AI Applications(KI - Künstliche Intelligenz: Vol. 38, No. 0, 2024) Steen, Alexander; Benzmüller, ChristophIn knowledge representation and reasoning, a key area in artificial intelligence research, non-classical logics play a prominent double role: firstly, non-classical logic languages allow for a precise and transparent encoding of domain specific knowledge. Secondly, as the logical languages are equipped with custom-tailored rules of logical inference, they make available a principled approach to derive new knowledge from previous information. In practice, the first aspect addresses data storage and retrieval, the second aspect the utilization of available information. This article briefly surveys contemporary challenges of NCL research in AI.
- ZeitschriftenartikelCLKR: Conditional Logic and Knowledge Representation(KI - Künstliche Intelligenz: Vol. 38, No. 0, 2024) Beierle, Christoph; Haldimann, Jonas; Schwarzer, LeonCLKR (Conditional Logic and Knowledge Representation) is an online repository of conditional logic resources for knowledge representation and reasoning. The question which entailments should follow from a conditional knowledge base consisting of a set of conditionals “ If A then usually B “ is central in logic-based AI. In order to support the practical side of this question, CLKR provides various collections of conditional knowledge bases and related resources. All knowledge bases available in CLKR can be processed directly with a corresponding reasoning system like InfOCF-Web. The sets of knowledge bases include examples as they are used in the literature for illustration, application knowledge bases from different domains, and systematically generated knowledge bases for evaluating implementations of nonmonotonic reasoning. A main emphasis of the current version of CLKR is on providing collections of knowledge bases in various normal forms that have been proposed for conditional knowledge bases, e.g., conditional normal form, antecedent normal form, and renaming normal form.
- ZeitschriftenartikelComputer-Verified Foundations of Metaphysics(KI - Künstliche Intelligenz: Vol. 38, No. 0, 2024) Kirchner, Daniel
- ZeitschriftenartikelEye of the Beholder(KI - Künstliche Intelligenz: Vol. 38, No. 0, 2024) Richter, Kai-Florian
- ZeitschriftenartikelLearning Normative Behaviour Through Automated Theorem Proving(KI - Künstliche Intelligenz: Vol. 38, No. 0, 2024) Neufeld, Emery A.Reinforcement learning (RL) is a powerful tool for teaching agents goal-directed behaviour in stochastic environments, and many proposed applications involve adopting societal roles which have ethical, legal, or social norms attached to them. Though multiple approaches exist for teaching RL agents norm-compliant behaviour, there are limitations on what normative systems they can accommodate. In this paper we analyse and improve the techniques proposed for use with the Normative Supervisor (Neufeld, et al., 2021)—a module which uses conclusions gleaned from a defeasible deontic logic theorem prover to restrict the behaviour of RL agents. First, we propose a supplementary technique we call violation counting to broaden the range of normative systems we can learn from, thus covering normative conflicts and contrary-to-duty norms. Additionally, we propose an algorithm for constructing a “normative filter”, a function that can be used to implement the addressed techniques without requiring the theorem prover to be run at each step during training or operation, significantly decreasing the overall computational overhead of using the normative supervisor. In order to demonstrate these contributions, we use a computer game-based case study, and thereafter discuss remaining problems to be solved in the conclusion.
- ZeitschriftenartikelModeling $$\mathscr {C}^{0}$$ C 0 Family Logics for Artificial Intelligence: Doxastic-Temporal Logics for Reasoning About Goals(KI - Künstliche Intelligenz: Vol. 38, No. 0, 2024) Oswald, James T.; Rozek, Brandon; Ferguson, Thomas M.We introduce the $$\mathscr {C}^{0}$$ C 0 family of logics, which include temporalized modal operators for belief and hyperintensional modal operators for obligations and goals. We motivate the $$\mathscr {C}^{0}$$ C 0 family as extended doxastic fragments of the $$\mathcal {DCEC}$$ DCEC family of logics, which are cognitive calculi designed for theory-of-mind reasoning among multiple artificial agents. In the literature, $$\mathcal {DCEC}$$ DCEC family logics are defined exclusively using proof-theoretic semantics. In this work we provide a model theory for the $$\mathscr {C}^{0}$$ C 0 family of logics which constitutes the first steps towards providing a model theory for the $$\mathcal {DCEC}$$ DCEC cognitive calculi family as a whole. We investigate the fragment relationships between both the $$\mathscr {C}^{0}$$ C 0 family and the $$\mathcal {DCEC}$$ DCEC family, produce a model theory for the $$\mathscr {C}^{0}$$ C 0 family and prove important results establishing completeness for all $$\mathscr {C}^{0}$$ C 0 family logics and establish soundness for $$\mathscr {C}^{0}$$ C 0 fragments without time.
- ZeitschriftenartikelNon-Classical Reasoning for Contemporary AI Applications(KI - Künstliche Intelligenz: Vol. 38, No. 0, 2024) Steen, Alexander; Benzmüller, Christoph
- ZeitschriftenartikelReport on “Axiomatizing Conditional Normative Reasoning”(KI - Künstliche Intelligenz: Vol. 38, No. 0, 2024) Parent, XavierThis is a report on the project “Axiomatizing Conditional Normative Reasoning” (ANCoR, M 3240-N) funded by the Austrian Science Fund (FWF). The project aims to deepen our understanding of conditional normative reasoning by providing an axiomatic study of it at the propositional but also first-order level. The focus is on a particular framework, the so-called preference-based logic for conditional obligation, whose main strength has to do with the treatment of contrary-to-duty reasoning and reasoning about exceptions. The project considers not only the meta-theory of this family of logics but also its mechanization.
- ZeitschriftenartikelSemantics of Belief Change Operators for Intelligent Agents(KI - Künstliche Intelligenz: Vol. 38, No. 0, 2024) Sauerwald, KaiThis paper summarises several contributions to the theory of belief change by the authors’ dissertation thesis. First, a relational characterization of belief revision for Tarskian logics is considered, encompassing first-order predicate logic, description logic, modal logics and many monotonic logics with model-theoretic semantics. Those logics where total preorders are the standard semantics for revision are characterized. The second contribution considered is a theory of belief revision that builds upon the idea that agents are limited in what the outcome of a revision is. Furthermore, advancements in principles for iterated belief contraction given in the thesis are outlined.