I have the same opinion on Rapid Miner. It seems that often times people look for the 'best' Open Source or commercial package. Often, I think this is the wrong approach. I have been more successful working with the most popular ones, especially with the open source packages. I think this is true because so much has already been developed that you can use as a base & build from. It cuts your development time down a lot. I don't think very many people are coming up with something that hasn't been done before in an open source project such as R. I am currently working on a project using SugarCRM (most popular open source web based CRM software) as a base & have yet to find anything that I need to do that hasn't already been done & made available (in some form) that I can find on the top page or two of google search results. Eric
Also this book appears to be finally being printed in jan... http://www.liaad.up.pt/~ltorgo/DataMiningWithR/ You can download an ancient copy of it from 2002... One case study is on the market. Will be interesting to see what other case studies are in it.