A SAT-Based Approach for Mining Association Rules

Authors: Abdelhamid Boudane, Said Jabbour, Lakhdar Sais, Yakoub Salhi

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

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on many datasets show that on both closed and indirect association rules mining tasks, our declarative approach achieves better performance than the state-of-the-art specialized techniques.
Researcher Affiliation Academia CRIL CNRS, Universit e d Artois Rue Jean Souvraz, SP-18 62307, Lens Cedex 3
Pseudocode No The paper contains mathematical formulas and logical expressions, but no explicitly labeled 'Pseudocode' or 'Algorithm' blocks are present.
Open Source Code No The paper does not state that the code for their proposed SAT-based approach is open-source or provide a link to a repository.
Open Datasets Yes Audiology (148, 216, 45%), Zoo-1 (36, 101, 44%), Tic-tac-toe (27, 958, 33%), Anneal (93, 812, 45%), Australian-credit (125, 653, 41%), German-credit (112, 1000, 34%), Heart-cleveland (95, 296, 47%), Hepatitis (68, 137, 50), Hypothyroid (88, 3247, 49%), Kr-vs-kp (73, 3196, 49%), Lymph (68, 148, 40%), Mushroom (119, 8124, 18%), Primary-tumor (31, 336, 48%), Soybean (50, 650, 32%), Splice-1 (287, 3190, 21%), Vote (48, 435, 33%)
Dataset Splits No For pure and closed association rules, the support is varied from 5% to 100% with an interval of size 5%. The confidence is varied in the same way. Then, for each data, a set of 400 configurations is generated. For indirect association rules, there are an additional parameter λ. The frequency and confidence are varied from 20% to 100% with an interval of size 20%. λ is varied from 10% to 100% with an interval of size 10%. This leads to 250 configurations for each data.
Hardware Specification Yes All the experiments were done on Intel Xeon quad-core machines with 32GB of RAM running at 2.66 Ghz.
Software Dependencies No The paper mentions using 'modern SAT solvers' and comparing against 'Coron' and 'SPMF' toolkits, but it does not provide specific version numbers for these software components.
Experiment Setup Yes For pure and closed association rules, the support is varied from 5% to 100% with an interval of size 5%. The confidence is varied in the same way. Then, for each data, a set of 400 configurations is generated. For indirect association rules, there are an additional parameter λ. The frequency and confidence are varied from 20% to 100% with an interval of size 20%. λ is varied from 10% to 100% with an interval of size 10%. This leads to 250 configurations for each data. ... For each instance, we fix the timeout to 15 minutes of CPU time.