Despite my very limited German (the presentation was held in German :-( ), I think I managed to convey the idea that as soon as somebody has data from his/her business, he/she can learn something.
I started with a few examples to make the definition of data mining from Fayyad, Piatetsky-Shapiro, and Smith (1996) more tangible. Some of those examples I have already reported in this blog in the past weeks/months.
The first example was the Dr. Semmelweis story (blog of 05/21/2010). I used this story to show that data mining is the "extraction of knowledge from data".
Then I told the auto insurance story (remember? Insurancies offer cheaper prices to women, blog of June 25th 2010). Here I wanted to show that data mining must be "potentially useful", that is it helps us to make more money.
Finally I told the story of the storcks and the babies (you can statistically prove that storcks bring babies) to show that data mining must extract knowledge that is actually "valid", that is it must correctly extrapolate the knowledge from the data. This story I have not told yet on this blog.
After that I went on with a few examples of very successful companies that owe their success to advanced techniques in data mining, like Netflix, eHarmony, Amazon, the Oakland A's, etc .... All pretty much handbook stories.
The audience was small but very active. Some people were definitely interested in the possibility of saving money or increasing profits by using analytics.
Let me know if you want a copy of the slides, by leaving a comment on this blog or by using the contact form!