On the Computation of Paracoherent Answer Sets

Authors: Giovanni Amendola, Carmine Dodaro, Wolfgang Faber, Nicola Leone, Francesco Ricca

AAAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental An empirical performance comparison among the new algorithms on benchmarks from ASP competitions is given as well. Finally, we report the results of an experimental activity conducted on benchmarks from ASP competitions (Calimeri et al. 2016), identifying the more efficient algorithm.
Researcher Affiliation Academia 1Department of Mathematics and Computer Science, University of Calabria, Italy {amendola,dodaro,leone,ricca}@mat.unical.it 2School of Computing and Engineering, University of Huddersfield, United Kingdom w.faber@hud.ac.uk
Pseudocode Yes Algorithm 1: Filtering Algorithm 2: Guess&Check Algorithm 3: Minimize Algorithm 4: Split
Open Source Code Yes WASP is an open-source ASP solver, winner of the latest ASP competition (Gebser, Maratea, and Ricca 2015), that we modified by implementing the algorithms presented in the previous section (the source can be downloaded at https://github.com/alviano/wasp).
Open Datasets Yes We use benchmark instances from the latest ASP competition (Gebser, Maratea, and Ricca 2015) collection. We consider all the incoherent instances that do not feature in the encoding neither aggregates, nor choice rules, nor weak constraints, since such features are not currently supported by the paracoherent semantics (Amendola et al. 2016a).
Dataset Splits No The paper refers to using 'benchmarks from ASP competitions' but does not explicitly state any specific training/validation/test splits, percentages, or sample counts within the paper.
Hardware Specification Yes Experiments were run on a Debian Linux system with 2.30GHz Intel Xeon E5-4610 v2 CPUs and 128GB of RAM.
Software Dependencies No The paper mentions 'a Java rewriter', 'state-of-the-art ASP solver WASP', and 'GRINGO (from http://potassco.sourceforge.net/)', but does not provide specific version numbers for any of these software components.
Experiment Setup Yes Execution time and memory were limited to 1200 seconds and 3 GB, respectively.