General Acyclicity and Cyclicity Notions for the Disjunctive Skolem Chase

Authors: Lukas Gerlach, David Carral

AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Our empirical evaluation shows that our novel notions are significantly more general than existing criteria. In Section 5, we empirically show that MFA and MFC are significantly less general than our novel conditions, which allow us to establish (non-)termination of many of the (non-deterministic) rule sets in our test suite.
Researcher Affiliation Academia Lukas Gerlach1, David Carral2 1 Knowledge-Based Systems Group, TU Dresden, Dresden, Germany 2 LIRMM, Inria, University of Montpellier, CNRS, Montpellier, France lukas.gerlach@tu-dresden.de, david.carral@inria.fr
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks.
Open Source Code Yes The tools, rule sets, and results of the evaluation are available online.8 Further information on the concrete steps to reproduce the evaluation steps is also provided there. 8https://doi.org/10.5281/zenodo.7375461 Gerlach and Carral
Open Datasets Yes The ontologies come from the Oxford Ontology Repository (OXFD),9 the dataset of the OWL Reasoner Evaluation 2015 (ORE15),10 and the Manchester OWL Corpus (MOWL).11 9https://www.cs.ox.ac.uk/isg/ontologies/ 10https://doi.org/10.5281/zenodo.18578 Parsia et al. 11https://doi.org/10.5281/zenodo.16708 Matentzoglu et al.
Dataset Splits No The paper evaluates properties of rule sets rather than training a machine learning model, and thus does not describe typical train/validation/test dataset splits.
Hardware Specification No The paper does not provide specific hardware details (like GPU/CPU models or memory) used for running the experiments.
Software Dependencies No The paper mentions obtaining rule sets from OWL ontologies and refers to 'tools' being available online, but it does not specify software dependencies (e.g., libraries, frameworks) with version numbers that would be needed to replicate the experimental setup.
Experiment Setup Yes To avoid an exponential number of checks, we consider a simplified version of DMFC in our implementation... We set a timeout of 30 minutes for each check... We consider an improvement of DMFA in our implementation... Definition 22. A rule set R is DMFAk for some k 1 if DMFA (R) does not feature any k-cyclic term; that is, a term with k + 1 nested occurrences of the same function symbol. Hence, we have decided to not publish results for k > 2 and believe that using δ-boundedness would not result in a big increase in performance in practice.