GEQCA: Generic Qualitative Constraint Acquisition

Authors: Mohamed-Bachir Belaid, Nassim Belmecheri, Arnaud Gotlieb, Nadjib Lazaar, Helge Spieker3690-3697

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

Reproducibility Variable Result LLM Response
Research Type Experimental In this section, we experimentally evaluate GEQCA, with time intervals as entities, on learning temporal constraints based on the Allen s interval algebra, where |Γ| = 13 (see Figure 1). The goal is 1) to compare our approach with LQCN (Mouhoub, Marri, and Alanazi 2018) 2) to evaluate our contributions for learning temporal networks and real-world scheduling problems.
Researcher Affiliation Academia 1Simula Research Laboratory, Oslo, Norway 2LITIO, University of Oran 1, Oran, Algeria 3LIRMM, University of Montpellier, CNRS, Montpellier, France
Pseudocode Yes Algorithm 1: GEQCA: Constraint Acquisition via Qualitative Queries.
Open Source Code Yes The code and full descriptions of each instance are publicly available at https://github.com/lirmm/Constraint Acquisition/tree/GEQCA.
Open Datasets Yes We use the publicly available RCPSP instances3 and we consider the structure of the problem with tasks duration, resource requirements, and resource capacities as background knowledge K, noted K1. Note 3: https://github.com/Mini Zinc/minizincbenchmarks/tree/master/rcpsp
Dataset Splits No The paper describes the generation of random problem instances and the use of RCPSP instances, but it does not specify explicit training, validation, or test dataset splits (e.g., percentages or sample counts).
Hardware Specification Yes All tests are run on an Intel core i7, 2.8GHz with RAM of 16GB.
Software Dependencies No GEQCA is implemented in Java and using Choco solver for the solve procedure. The paper does not provide specific version numbers for Java or the Choco solver.
Experiment Setup Yes In Table 2, we report the user effort on 5 scheduling instances using GEQCA with PATH heuristic, cutoff = 3, 600s and K { , K1, K1 K2}. Hence, we conduct a new experiment by setting the cutoff time to 2sec.