ECML PKDD 2008

From Yale to RapidMiner: A Guide on How to Live on Free Data Mining Software

Back in 2001, a small team of ambitious scientists around Prof. Dr. K. Morik at the University of Dortmund was tired of adapting and re-defining a bunch of Perl scripts for each data analysis task anew (the elder of us will probably remember). Searching for an alternative solution, they identified some shortcomings of the existing data mining suites and started the development of a free open-source solution. Filled with modesty, they called this software “Yale” which was an acronym for “Yet Another Learning Environment” – yet another one beside those big players like SPSS, SAS, and of course Weka. Seven years later, the software has matured and so have its now world-wide more than 30 developers: Yale had to be renamed to “RapidMiner”, a spin-off company “Rapid-I” was founded not only maintaining the software but also providing all types of services around RapidMiner, and first real-world experiences had to be gained. This talk will include several real-world stories showing - the gap between scientific data mining and the desires of people in the real-world, - the process of changing the scientist’s way of thinking to fill this gap, - and: how on earth can people live on selling something that is free?