%0 Journal Article %1 padillo2017exhaustive %A Padillo, F. %A Luna, J. M. %A Ventura, S. %D 2017 %J Progress in Artificial Intelligence %K distributed emm mapreduce parallel spark subgroup subgroups %P 1--14 %R 10.1007/s13748-017-0112-x %T Exhaustive search algorithms to mine subgroups on Big Data using Apache Spark %U http://dx.doi.org/10.1007/s13748-017-0112-x %X Subgroup discovery is a well-known technique for the extraction of patterns, with respect to a variable of interest in the data. However, the explosion in data gathering has hampered the performance of traditional algorithms to discover interesting relationships between different objects in a set with respect to a specific property which is of interest to the user. In this regard, our goal is to propose a set of efficient techniques to mine subgroups on Big Data by means of Apache Spark. On this matter, AprioriK-SD-OE and PFP-SD-OE are proposed as fast exhaustive search algorithms to discover subgroups on Big Data using Apache Spark. The experimental study includes more than 70 datasets considering search spaces bigger than \$\$10^\15\\$\$ 10 15 subgroups. The scalability of our proposals are analyzed by considering datasets with 200 million of instances demonstrating the usefulness of using Spark to tackle Big Data.

@article{padillo2017exhaustive, abstract = {Subgroup discovery is a well-known technique for the extraction of patterns, with respect to a variable of interest in the data. However, the explosion in data gathering has hampered the performance of traditional algorithms to discover interesting relationships between different objects in a set with respect to a specific property which is of interest to the user. In this regard, our goal is to propose a set of efficient techniques to mine subgroups on Big Data by means of Apache Spark. On this matter, AprioriK-SD-OE and PFP-SD-OE are proposed as fast exhaustive search algorithms to discover subgroups on Big Data using Apache Spark. The experimental study includes more than 70 datasets considering search spaces bigger than {\$}{\$}10^{\{}15{\}}{\$}{\$} 10 15 subgroups. The scalability of our proposals are analyzed by considering datasets with 200 million of instances demonstrating the usefulness of using Spark to tackle Big Data.}, added-at = {2017-02-01T09:17:54.000+0100}, author = {Padillo, F. and Luna, J. M. and Ventura, S.}, biburl = {https://www.bibsonomy.org/bibtex/298dc9843d08ee2d2e0141617ddb2c33d/becker}, description = {Exhaustive search algorithms to mine subgroups on Big Data using Apache Spark | SpringerLink}, doi = {10.1007/s13748-017-0112-x}, interhash = {e2d9e11931a577ba376ca8b91c263435}, intrahash = {98dc9843d08ee2d2e0141617ddb2c33d}, issn = {2192-6360}, journal = {Progress in Artificial Intelligence}, keywords = {distributed emm mapreduce parallel spark subgroup subgroups}, pages = {1--14}, timestamp = {2017-02-01T09:17:54.000+0100}, title = {Exhaustive search algorithms to mine subgroups on Big Data using Apache Spark}, url = {http://dx.doi.org/10.1007/s13748-017-0112-x}, year = 2017 }