Poll: Current

Your data mining experience (in years): [264 votes total] none (45) 17% 1-2 (61) 23% 3-5 (54) 20% 6-10 (67) 25% 11-20 (33) 13% 21+ (4) 2%

Dr S M Patel, data mining in educational field
I have been working in academic field from past 22 years and doing lot of data analysis work using spss packages or tool developed as in-house for the different requirement. We have achieved very good results as a part of this continuous work. I suggest each company/ organization must have a seperate (small) department for data mining to work for meeting the need of the future.

Nikolay, 21+20 years of experience
40 years back we have understood, that tasks, in which we were engaged (pattern recognition, automatic classification, a choice of informative attributes, filling gaps and detection of mistakes in data tables, forecasting) are based on a fundamental problem of discovering of empirical regularities in data file. Thus the special attention was given such characteristics of many real tasks as different feature scales, presence of blanks and noise, small volume of sample, absence of the information on the law of distribution, and so forth. After 20 years we have found out that this direction began to refer to Data Mining. So we vote for a position 21+20.

Bill, Years of data mining experience
It would be interesting to see this question broken out between applied vs. research experience. While a good understanding of the theoretical underpinnings of different data mining algorithms is a must in order to properly use them, there is a tremendous difference in required skillsets between applying this knowledge in a research setting and being asked to provide key insights or a good predictive model in 5 to 7 days from a dirty dataset as often happens in rapid cycle business settings. The tradeoff between the two is the elegance of the solution vs. choosing methods that are scalable and relatively robust in application.

Karl, What to count
I'd say if you were writing about data mining, or substantially about it, then time writing, and indeed doing DM research for, a PhD definitely counts. If you ask me it's likely to have been far more instructive experience than just turning the handle on some software packages. And I've no doubt some people count that as doing DM.