Simulating Sets in Answer Set Programming

Authors: Sarah Alice Gaggl, Philipp Hanisch, Markus Krötzsch

IJCAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Our evaluation considers problems of ontological reasoning that are not in scope for traditional ASP (unless EXPTIME = ΠP 2), and we find that our new existential-rule grounding performs well in comparison with native implementations of set terms in ASP.
Researcher Affiliation Academia Sarah Alice Gaggl1 , Philipp Hanisch2 and Markus Kr otzsch2 1Logic Programming and Argumentation Group, TU Dresden, Germany 2Knowledge-Based Systems Group, TU Dresden, Germany {sarah.gaggl, philipp.hanisch1, markus.kroetzsch}@tu-dresden.de
Pseudocode No The paper presents logical rules (e.g., rules (7)-(14), (17)-(20)) and transformations, but these are not explicitly labeled as 'Pseudocode' or 'Algorithm' blocks.
Open Source Code Yes Full proofs, datasets, and source code for our prototype implementations can be accessed through the online repository https://github.com/knowsys/eval-2022-IJCAI-asp-with-sets.
Open Datasets Yes Full proofs, datasets, and source code for our prototype implementations can be accessed through the online repository https://github.com/knowsys/eval-2022-IJCAI-asp-with-sets.
Dataset Splits No The paper describes using entire ontologies for evaluation and does not mention explicit training, validation, or test dataset splits in the typical sense for machine learning models. The tasks involve inference on complete knowledge bases.
Hardware Specification Yes Our evaluation computer is a mid-end server (Debian Linux 9.13; Intel Xeon CPU E5-2637v4@3.50GHz; 384GB RAM DDR4; 960GB SSD), though most experiments did not use more than 8GB of RAM.
Software Dependencies Yes The first approach (LAZY) uses the transformation dlpf(P) and the lazy ASP solver Alpha3 [Weinzierl, 2017]. The second approach (EXRULES) implements grounding with existential rules using the Rulewerk library with the VLog reasoner4 [Carral et al., 2019a] to produce a grounded file in aspif format, to which we apply the ASP solver clasp v3.2.1 [Gebser et al., 2007]. 3https://github.com/alpha-asp/Alpha, master, 29 Aug 2021 4https://github.com/karmaresearch/vlog, master, 30 Aug 2021
Experiment Setup Yes A timeout of 15min was used in all experiments.