Advances in frequent itemset mining implementations: report on FIMI'03

1. WHY ORGANIZE FIMI? Since the introduction of association rule mining in 1993 by Agrawal Imielinski and Swami [3], the frequent itemset mining (FIM) tasks have received a great deal of attention. Within the last decade, a phenomenal number of algorithms have been developed for mining all [3; 5; 20; 4; 27; 24; 29; 34; 10; 22; 19; 32], closed [25; 6; 12; 26; 30; 23; 31; 28; 33] and maximal frequent itemsets [18; 21; 7; 2; 1; 11; 36; 16; 17]. Every new paper claims to run faster than previously existing algorithms, based on their experimental testing, which is oftentimes quite limited in scope, since many of the original algorithms are not available due to intellectual property and copyright issues. Zheng, Kohavi and Mason [35] observed that the performance of several of these algorithms is not always as claimed by its authors, when tested on some different datasets. Also, from personal experience, we noticed that even different implementations of the same algorithm could behave quite differently for various datasets and parameters. Given this proliferation of FIM algorithms, and sometimes contradictory claims, there is a pressing need to benchmark, characterize and understand the algorithmic performance space. We would like to understand why and under what conditions one algorithm outperforms another. This means testing the methods for a wide variety of parameters, and on different datasets spanning dense and sparse, real and synthetic, small and large, and so on. Given the experimental, algorithmic nature of FIM (and most of data mining in general), it is crucial that other researchers be able to independently verify the claims made in a new paper. Unfortunately, the FIM community (with few exceptions) has a very poor track record in this regard. Many new algorithms are not available even as an executable, let alone the source code. How many times have we heard “this is proprietary software, and not available.” This is not the way other sciences work. Independent verifiability is the hallmark of sciences like physics, chemistry, biology, and so on. One may argue, that the nature of research is different, they have detailed experimental procedure that can be replicated, while we have algorithms, and there is more than one way to code an algorithm. However, a good example to emulate is the bioinformatics community. They have espoused the open-source paradigm with more alacrity than we have. It is quite common for journals and conferences in bioinformatics to require that software be available. For example, here is a direct quote from the journal Bioinformatics (http://bioinformatics.oupjournals.org/):