Tabling as a Library with Delimited Control

Authors: Benoit Desouter, Marko van Dooren, Tom Schrijvers, Alexander Vandenbroucke

IJCAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Table 1 summarizes the results (in ms) obtained on a Dell Power Edge R410 server (2.4 GHz, 32 GB RAM) running Debian 7.6. ... We consider the performance results of our implementation very reasonable, especially if we take into account the stark contrast between our concise and high-level implementation and the complex integration in other systems.
Researcher Affiliation Academia Benoit Desouter, Marko van Dooren Ghent University, Belgium... Tom Schrijvers, Alexander Vandenbroucke KU Leuven, Belgium
Pseudocode Yes Figure 1: Delimited execution. (Contains Prolog code snippet which explains the algorithm)
Open Source Code Yes The description and code of the benchmarks can be found at http://users.ugent.be/~bdsouter/tabling/.
Open Datasets No The paper uses standard benchmarks (e.g., 'recognize', 'n-reverse', 'shuttle'), but does not provide explicit links, DOIs, or citations for the datasets used by these benchmarks. It only links to the benchmark *code* at http://users.ugent.be/~bdsouter/tabling/.
Dataset Splits No The paper evaluates the system on benchmarks but does not mention specific training, validation, or test dataset splits.
Hardware Specification Yes Table 1 summarizes the results (in ms) obtained on a Dell Power Edge R410 server (2.4 GHz, 32 GB RAM) running Debian 7.6.
Software Dependencies Yes we compare our implementation in h Prolog 3.2.38 against XSB 3.4.0 [Swift and Warren, 2012], B-Prolog 8.1 [Zhou, 2012], Yap 6.3.4 [Santos Costa et al., 2012] and Ciao 1.15-2731-g3749edd [Hermenegildo et al., 2012]... running Debian 7.6.
Experiment Setup No The paper describes the general environment and versions of Prolog systems used but does not provide specific hyperparameters or detailed system-level training settings beyond the general setup of running benchmarks.