Towards easy prototyping of pattern mining problems

National audienceThe discovery of frequent patterns is a famous problemin data mining. While plenty of algorithms have been pro-posed during the last decade, only a few contributions havetried to understand the influence of datasets on the algo-rithms behavior. Being able to explain why certain algo-rithms are likely to perform very well or very poorly onsome datasets is still an open question.In this setting, we describe a thorough experimentalstudy of datasets with respect to frequent itemsets. We studythe distribution of frequent itemsets with respect to itemsetssize together with the distribution of three concise represen-tations: frequent closed, frequent free and frequent essentialitemsets. For each of them, we also study the distribution oftheir positive and negative borders whenever possible.From this analysis, we exhibit a new characterization ofdatasets and some invariants allowing to better predict thebehavior of well known algorithms.The main perspective of this work is to devise adaptivealgorithms with respect to dataset characteristics.The discovery of frequent patterns is a famous problemin data mining. While plenty of algorithms have been pro-posed during the last decade, only a few contributions havetried to understand the influence of datasets on the algo-rithms behavior. Being able to explain why certain algo-rithms are likely to perform very well or very poorly onsome datasets is still an open question.In this setting, we describe a thorough experimentalstudy of datasets with respect to frequent itemsets. We studythe distribution of frequent itemsets with respect to itemsetssize together with the distribution of three concise represen-tations: frequent closed, frequent free and frequent essentialitemsets. For each of them, we also study the distribution oftheir positive and negative borders whenever possible.From this analysis, we exhibit a new characterization ofdatasets and some invariants allowing to better predict thebehavior of well known algorithms.The main perspective of this work is to devise adaptivealgorithms with respect to dataset characteristics