The Sixth Answer Set Programming Competition
Authors: Martin Gebser, Marco Maratea, Francesco Ricca
JAIR 2017 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In this paper, we report about the design and results of the Sixth ASP Competition, which was jointly organized by the University of Calabria (Italy), Aalto University (Finland), and the University of Genoa (Italy), in affiliation with the 13th International Conference on Logic Programming and Non-Monotonic Reasoning. ... We present the results of the Sixth ASP Competition and analyze system performance with respect to language features, computational tasks, reference systems from the previous edition, effect of modeling, runtime limits, and sequential versus parallel processing. |
| Researcher Affiliation | Academia | Martin Gebser EMAIL If ICS, University of Potsdam, August-Bebel-Str. 89, 14482 Potsdam, Germany; Marco Maratea EMAIL DIBRIS, University of Genoa, Viale F. Causa 15, 16145 Genova, Italy; Francesco Ricca EMAIL De Ma CS, University of Calabria, Viale P. Bucci, Cubo 31B, 87036 Rende, Italy |
| Pseudocode | No | The paper provides examples of Answer Set Programming (ASP) rules and encodings, such as those in Section 2.4 describing the Traveling Salesperson Problem or in Section 5.2 for new domains like Combined Configuration. These are code snippets illustrating ASP syntax and problem formulations, but they are not presented as structured pseudocode or algorithm blocks describing the methodology of a system developed by the authors. |
| Open Source Code | No | The paper reports on the design and results of the Sixth ASP Competition, detailing the various participant systems (solvers) and their approaches. While it mentions 'tools to reproduce the instance selection are available on-line (ASP-Comp, 2015)', it does not provide source code for the methodologies of the participant solvers described in the paper or for any novel methodology developed by the authors of this paper. The participant systems themselves (e.g., CLASP, WASP, ME-ASP) are third-party tools. |
| Open Datasets | Yes | Industrial instances, taken from the 2014 Max SAT Evaluation (Max SAT-Comp, 2014), are represented by facts and encoded as an Optimization problem. ... The resulting instances run in the competition and tools to reproduce the instance selection are available on-line (ASP-Comp, 2015). |
| Dataset Splits | No | For each of the 28 domains used for ranking participant systems, the selection task was to pick 20 instances to run in the competition. As mentioned above, non-groundable and very easy instances were excluded, so that the selection is made among easy, medium, hard, and too hard instances. ... The goal of an even division of instances to pick from C can be expressed as follows, where s ∈ {sat, unsat} ... The paper describes a strategy for selecting competition instances based on empirical hardness and satisfiability categories. This is a selection process for competition benchmarks, not a training/test/validation split in the context of model evaluation or machine learning experiments. |
| Hardware Specification | Yes | The competition is run on a Debian Linux server (64bit kernel), equipped with 2.30GHz Intel Xeon E5-4610 v2 Processors and 128GB RAM. |
| Software Dependencies | No | The paper refers to several software systems used in the competition or as part of participant systems, such as 'GRINGO for grounding', 'CLASP', 'WASP', 'DLV', 'CPLEX', 'LINGELING', and 'SAT4J (Le Berre & Parrain, 2010)'. While it mentions 'CLASP version of 2014' and 'WASP-1.5' in the context of reference systems from the previous edition, it does not consistently provide specific version numbers for the key software components used by the participant systems in the Sixth ASP Competition, which are necessary for full reproducibility. |
| Experiment Setup | Yes | Time and memory for each run are limited to 20 minutes wall-clock time and 12GB, respectively. Participant systems can exploit up to eight cores (16 virtual CPUs, given that Intel Hyperthreading technology is enabled) in the MP category, while the execution is constrained to one core in the SP category. ... Pragmatically, we picked the instances provided in the first answer set computed by CLASP, using its options rand-freq, sign-def, and seed for guaranteeing reproducible randomization. In particular, we fixed the random seed to the concatenation of winning numbers in the Euro Millions lottery of 23rd June 2015. |