Efficient HEX-Program Evaluation Based on Unfounded Sets
Authors: T. Eiter, M. Fink, T. Krennwallner, C. Redl, P. Schüller
JAIR 2014 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | An experimental evaluation on advanced reasoning applications shows that unfounded sets checking combined with learning methods of Eiter et al. (2012a) improves HEX-program evaluation considerably, sometimes drastically. |
| Researcher Affiliation | Academia | Thomas Eiter EMAIL Michael Fink EMAIL Thomas Krennwallner EMAIL Christoph Redl EMAIL Institut f ur Informationssysteme Technische Universit at Wien Favoritenstraße 9-11, A-1040 Vienna, Austria Peter Sch uller EMAIL Faculty of Engineering and Natural Sciences Sabanci University Orhanli, Tuzla, 34956 Istanbul, Turkey |
| Pseudocode | No | The paper describes formal encodings and logical rules, such as nogood sets, for unfounded set detection, but it does not contain clearly labeled pseudocode or algorithm blocks with structured steps. |
| Open Source Code | Yes | Benchmark encodings and HEX-plugins are publicly available at https:/github.com/ hexhex/benchmarks. |
| Open Datasets | Yes | We use random instances of different MCS topologies, i.e., connection graphs of contexts, created with our MCS benchmark generator.3 Described at http:/www.kr.tuwien.ac.at/research/systems/dmcs/experiments.html, online available at https:/dmcs. svn.sourceforge.net/svnroot/dmcs/dmcs/trunk |
| Dataset Splits | No | The paper describes generating 'random instances' for benchmarks and uses different numbers of arguments or contexts, but it does not specify how these instances are split into training, validation, or test sets for reproduction. |
| Hardware Specification | Yes | We evaluated the implementation on a Linux server with two 12-core AMD 6176 SE CPUs with 128GB RAM running DLVHEX version 2.3.0. |
| Software Dependencies | Yes | For implementing our technique, we integrated CLASP into our prototype system DLVHEX; we use CLASP as an ASP solver for computing compatible sets and as a SAT solver for solving the nogood set of the UFS check. ... running DLVHEX version 2.3.0. |
| Experiment Setup | Yes | The timeout was uniformly set to 300 seconds for each instance; for each parameter value, the average runtime over all instances is printed where timeouts, whose number is shown in parentheses, are fully taken into account. ... In each test run the CPU usage was limited to two CPU cores, running a Condor load distribution system which ensures robust runtimes (i.e., multiple runs of the same instance have negligible deviations). |