Real example

MataDiner

Hi,

I am new to Data Mining. I looked at at the examples within the rapid miner tool and they all make sense. to me. What I would like is a more "realistic" example. In the built-in examples everything is very obvious, but with real data it might not be like that.

IngoRM

Hi,

the problem with real examples is: nobody is going to give you your data (at least not from commercial settings but even in science people are sometimes reluctant to do so - at least not before the corresponding paper was published). So it is hard to get a full real-live process including all specific details, the data etc. And you are right: the sample processes delivered with RapidMiner are rather simple compared to the processes we use, for example, for our customers. They are concentrating on one detail only (people should learn the concepts from them) and not on a full knowledge discovery process.

However, here is something which is more similar to a real-world example. On the feature list of the RapidMiner Enterprise Edition is also a full analysis process for a report (but again: it is rather small compared to most other real-life processes we work on - however, better than nothing):

http://rapid-i.com/content/view/123/141/

Just click on "Reporting Engine" and in the pop-up window select "Process for creating the HTML report (XML)". Then you will see a complete reporting process (it will not work with the community edition since some additional reporting operators are necessary which are only part of the Enterprise Editions). The "box view" in the same popup window will also give you a more visual idea of how this looks like.


Here is my suggestion: why are people not posting their processes (including even their data if possible and not too large) here or in our Wiki? So others could get new ideas of what can be done and how.

Cheers,
Ingo