Debugging Classical Ontologies Using Defeasible Reasoning Tools

A successful application of ontologies relies on representing as much accurate and relevant domain knowledge as possible, while maintaining logical consistency. As the successful implementation of a real-world ontology is likely to contain many concepts and intricate relationships between the concepts, it is necessary to follow a methodology for debugging and refining the ontology. Many ontology debugging approaches have been developed to help the knowledge engineer pinpoint the cause of logical inconsistencies and rectify them in a strategic way. We show that existing debugging approaches can lead to unintuitive results, which may lead the knowledge engineer to opt for deleting potentially crucial and nuanced knowledge. We provide a methodological and design foundation for weakening faulty axioms in a strategic way using defeasible reasoning tools. Our methodology draws from Rodler’s interactive ontology debugging approach and extends this approach by creating a methodology to systematically find conflict resolution recommendations. Importantly, our goal is not to convert a classical ontology to a defeasible ontology. Rather, we use the definition of exceptionality of a concept, which is central to the semantics of defeasible description logics, and the associated algorithm to determine the extent of a concept’s exceptionality (their ranking); then, starting with the statements containing the most general concepts (the least exceptional concepts) weakened versions of the original statements are constructed; this is done until all inconsistencies have been resolved.

Argumentation and Defeasible Reasoning in the Law

Different formalisms for defeasible reasoning have been used to represent knowledge and reason in the legal field. In this work, we provide an overview of the following logic-based approaches to defeasible reasoning: defeasible logic, Answer Set Programming, ABA+, ASPIC+, and DeLP. We compare features of these approaches under three perspectives: the logical model (knowledge representation), the method (computational mechanisms), and the technology (available software resources). On top of that, two real examples in the legal domain are designed and implemented in ASPIC+ to showcase the benefit of an argumentation approach in real-world domains. The CrossJustice and Interlex projects are taken as a testbed, and experiments are conducted with the Arg2P technology.

The Source Conflict Cost Criterion: A Pragmatic Approach to Floating Conclusions

<p>Our research examines the problem of multiple lines of reasoning reaching the same conclusion, but only through different and unrelated arguments. In the context of non-monotonic logic, these types of conclusions are referred to as floating conclusions. The field of defeasible reasoning is divided between those who claim that floating conclusions ought not to be accepted through a prudent or skeptical point of view, whereas others argue that they are good enough conclusions to be admitted even from a conservative or skeptical standard. We approach the problem of floating conclusions through the formal framework of Inheritance Networks. These networks provide the simplest and most straightforward gateway into the technical aspects surrounding floating conclusions in the context of non-monotonic logic and defeasible reasoning. To address the problem of floating conclusions, we construct a unifying framework of analysis, namely, the Source Conflict Cost Criterion (SCCC), that contains two basic elements: source conflict and cost. Both elements are simplified through a binary model, through which we provide a comprehensive understanding of the floating conclusions as well as the problematic nature of the debate surrounding this type of inferences. The SCCC addresses three key objectives: (a) the assessment of floating conclusions and the debate surrounding its epistemological dimension, (b) the construction of a general and unified framework of analysis for floating conclusions, and (c) the specification of the normative conditions for the admission of floating conclusions as skeptically acceptable information.</p>

The Source Conflict Cost Criterion: A Pragmatic Approach to Floating Conclusions

<p>Our research examines the problem of multiple lines of reasoning reaching the same conclusion, but only through different and unrelated arguments. In the context of non-monotonic logic, these types of conclusions are referred to as floating conclusions. The field of defeasible reasoning is divided between those who claim that floating conclusions ought not to be accepted through a prudent or skeptical point of view, whereas others argue that they are good enough conclusions to be admitted even from a conservative or skeptical standard. We approach the problem of floating conclusions through the formal framework of Inheritance Networks. These networks provide the simplest and most straightforward gateway into the technical aspects surrounding floating conclusions in the context of non-monotonic logic and defeasible reasoning. To address the problem of floating conclusions, we construct a unifying framework of analysis, namely, the Source Conflict Cost Criterion (SCCC), that contains two basic elements: source conflict and cost. Both elements are simplified through a binary model, through which we provide a comprehensive understanding of the floating conclusions as well as the problematic nature of the debate surrounding this type of inferences. The SCCC addresses three key objectives: (a) the assessment of floating conclusions and the debate surrounding its epistemological dimension, (b) the construction of a general and unified framework of analysis for floating conclusions, and (c) the specification of the normative conditions for the admission of floating conclusions as skeptically acceptable information.</p>

Defeasible Reasoning via Datalog¬

We address the problem of compiling defeasible theories to Datalog¬ programs. We prove the correctness of this compilation, for the defeasible logic DL(∂||), but the techniques we use apply to many other defeasible logics. Structural properties of DL(∂||) are identified that support efficient implementation and/or approximation of the conclusions of defeasible theories in the logic, compared with other defeasible logics. We also use previously well-studied structural properties of logic programs to adapt to incomplete Datalog¬ implementations.

Evidence Graphs: Supporting Transparent and FAIR Computation, with Defeasible Reasoning on Data, Methods and Results

Introduction: Transparency of computation is a requirement for assessing the validity of computed results and research claims based upon them; and it is essential for access to, assessment, and reuse of computational components. These components may be subject to methodological or other challenges over time. While reference to archived software and/or data is increasingly common in publications, a single machine-interpretable, integrative representation of how results were derived, that supports defeasible reasoning, has been absent. Methods: We developed the Evidence Graph Ontology, EVI, in OWL 2, with a set of inference rules, to provide deep representations of supporting and challenging evidence for computations, services, software, data, and results, across arbitrarily deep networks of computations, in connected or fully distinct processes. EVI integrates FAIR practices on data and software, with important concepts from provenance models, and argumentation theory. It extends PROV for additional expressiveness, with support for defeasible reasoning. EVI treats any com- putational result or component of evidence as a defeasible assertion, supported by a DAG of the computations, software, data, and agents that produced it. Results: We have successfully deployed EVI for very-large-scale predictive analytics on clinical time-series data. Every result may reference its own evidence graph as metadata, which can be extended when subsequent computations are executed. Discussion: Evidence graphs support transparency and defeasible reasoning on results. They are first-class computational objects, and reference the datasets and software from which they are derived. They support fully transparent computation, with challenge and support propagation. The EVI approach may be extended to include instruments, animal models, and critical experimental reagents.