Looking for (academic) references I can cite to support my use of RapidMiner

XAmsterdamX

I'm a new user of RapidMiner, only just scratching the surface of its capabilities but already impressed. I have used it recently to perform a cluster analysis on a purchase history database. I am currently writing a paper to publish the results in a Marketing journal.

My question:
Do any of you know of academic research that has used RapidMiner in the past? I have find a few using Google Scholar, but none have made an effort to support why they chose RapidMiner, or to prove its robustness / accuracy.

Alternatively, are there other sources (other than the Rapid-I website) that I can use to support my case?

Any suggestions are welcome. I will also gladly share my own paper as a reference for others once it's been accepted to the journal (which may take a while).

awchisholm

Hello

I used RapidMiner extensively in my Master's dissertation.

http://www.dataminingmasters.com/uploads/studentProjects/AChisholm2012ZipfsLawAuthorAttribution.pdf

regards

Andrew

XAmsterdamX

Thanks!

I had a quick look at your dissertation. Unfortunately you don't really address why you chose RapidMiner, at least not in a way that I can quote. I'm also not sure if I can cite a dissertation given the journal I am submitting my paper to, since it's not peer-reviewed.

I'll probably just have another look at Google Scholar; there are 1830 results to the search term 'rapidminer', there must be something useful there.

I'm not sure if Rapid-i employees read this forum, but in case they do: it may be a good idea to start collecting publications that used RapidMiner and show them on your website, since it gives the program credibility in academic research. You may even want to offer incentives to authors who publish research that used RapidMiner in peer-reviewed journals. For example, Leximancer (text analysis software) offers gives one-year licenses to authors who used it to publish a paper.

MariusHelf

Actually we do read the forums

Apart from other applications, RapidMiner is used in the astrophysics projects MAGIC and IceCube, probably you can find references to that via Google. Actually most papers won't state explicitly though why they chose RapidMiner, because the main focus of scientific work is based on the results and algorithms achieved and used. So the reasons will be rather implicit, such as stating that someone used RapidMiner, with this and that algorithm, and extended it to include it his own work etc... so the benefits are stated only implicitly (feature rich, easy to extend, ...).

You are right, collecting related work may be a good idea. Giving free licences to authors though is of limited use, since we distribute most of our software under the AGPL anyways

Best regards,
Marius

RWingerter

XAmsterdamX wrote:
... none have made an effort to support why they chose RapidMiner, or to prove its robustness / accuracy.

This market survey might be useful:

http://www.rexeranalytics.com/Data-Miner-Survey-Results-2011.html

Roland

RWingerter

Check out RCOMM 2010 Proceedings in the Rapid-I shop. It's a freedownload. The article "Pattern Recognition Engineering" includes an evaluation of RapidMiner and four other software packages.

Roland