On the Enumeration of Association Rules: A Decomposition-based Approach

Authors: Yacine Izza, Said Jabbour, Badran Raddaoui, Abdelahmid Boudane

IJCAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Finally, an experimental evaluation shows that our method is fast and scale well compared with the existing CP approach even in the sequential case, while significantly reducing the gap with the best state-of-the-art specialized algorithm. We now present the experiments carried out to evaluate the performance of our approach. In particular, we study the running time for discovering ARs and MNRs in sequential and parallel setting.
Researcher Affiliation Academia 1ANITI, Toulouse, France 2CRIL & CNRS, Universit e d Artois, Lens, France 3SAMOVAR, T el ecom Sud Paris, Institut Polytechnique de Paris, France 4LMTO, Ecole Militaire Polytechnique, Algeria
Pseudocode Yes Algorithm 1: Decomposition-based ARs Mining
Open Source Code Yes All experiments are available at: https://github.com/crillab/satar-xp.
Open Datasets Yes We consider the datasets used in [Belaid et al., 2019] coming from FIMI1 and CP4IM2 repositories. 1http://fimi.ua.ac.be/data/ 2http://dtai.cs.kuleuven.be/CP4IM/datasets/
Dataset Splits No The paper does not explicitly provide specific details on how the datasets were split into training, validation, or test sets, nor does it refer to predefined splits with citations.
Hardware Specification Yes Our experiments were performed on a machine with Intel Xeon quad-core processors with 32GB of RAM running at 2.66 GHz on Linux Cent OS.
Software Dependencies No The paper mentions that Algorithm 1 is implemented in C++ top-on the SAT solver Mini SAT [E en and S orensson, 2002] and uses the API Open MP. However, it does not provide specific version numbers for Mini SAT, Open MP, or any other software dependencies, which are necessary for full reproducibility.
Experiment Setup Yes The minimum confidence threshold β is fixed to 75%3 and different minimum support values are chosen w.r.t. the size of each dataset. Time-out was set to 1800 seconds and memory-out to 10 GB in all runs. The partition is performed by considering the items frequencies in increasing order.