Interesting discussion on our LinkedIn group, with 12 replies so far.
What data-mining software would you recommend? I'm using test item responses along with various demographic and longitudinal vars to help schools identify and group students for extra help.
We are using Excel and SAS currently, but could transform data for input into other software. We are especially interested in price and length of training needed to properly interpret the results.
R is free: www.r-project.org
, but takes a bit of time to get used to. If you've used SAS, you can probably use R with a little effort.
By now, everyone is sick of hearing me say how much I love R.
I can recommend Weka, RapidMiner and SAS Enterprise Miner. I think SAS EM is the best solution if you do not have any budget issues. Otherwise, start with Weka - easy to install, use, txt files can be imported, good visualization, a lot of DM methods. It also depends what you would like to do - even in SAS base you can create powerful models. There are also some add-ins for excel, able to create decision trees or even neural networks (but I have never tried these add-ins). Once I also used LispMiner, which was designed to mine for patterns in the form of certain type of generalized association rules ( lispminer.vse.cz ).
If it's improved visualization of your data, I'd recommend Tableau Software.
We typically take POS and time series data out of Oracle (using SAS) and down to the desktop by importing csv's or Access tables into Tableau. It's very inexpensive and very easy to use.
You can get a full version to try through the web (free download for 15 days).
For full discussion with links and authors, click here