|
For comparison, here are the results from 2007 KDnuggets Poll on Data Mining Software Popularity.
This year there were stronger poll verification measures, which eliminated some (but probably not all) biases due to over-enthusiastic voting of some vendors and users.
Among voters from US, the top choice was Salford.
Comments
Tim Manns, Real-world scaleability
I don't have big exposure of all these tools (my experience is limited to the major commerical tools), so am not certain that many of them are scaleable (or not).
In my role as a data miner for a telco I process many millions of rows, with hundreds of columns, and sample to many thousand when building predictive models. An easy UI tool that integrates with the data warehouse is a must, and for this reason I consider in-database mining and connectivity with the data warehouse a necessary requirement for a true data mining tool. - assuming your definition of data mining is (roughly) "to process large amounts of data and indentify useful actionable information".
I realise the world of data mining is varied, and db connectivity or scaleable performance is not always a consideration in these tools. You should too :)
Janaki Gopalan, DM tools
I have been using WEKA and I will be starting to use SPSS next month. In this poll, looks like WEKA is relatively popular free ware tool (6% users, after Rapid Miner (12%) and SPSS (9%). If you have used industry specific tools, can you tell me what is the advantage of them over free ware?