described as “unsupervised” or “supervised” in data mining?

Answer 1

"Supervised learning" is learning where the "right answer" is known. Examples would include "classification", in which the outcome is a category (i.e. predict whether a medical patient has disease X or not) and "regression", in which the outcome is numeric (for instance, predicting tomorrow's stock price of IBM).

"Unsupervised learning" is learning where there is no given "right answer" to use as a reference during learning. Most of the time, this term is used to mean "clustering", in which similar items are grouped together. Note the difference between clustering and classification: two perfectly valid clusterings of a set of items may yield vastly different results. Cars, for instance, might be assigne to clusters based on size, performance, efficiency or even color!

Answer 2

data mining is a computer term, has nothing to do with rocks and geology. the term has to do with a program that will look at database files and try and extract useful correlations that were not evident when the data was first collected. some enterprises run datamining on everything generated within the company, automated mindless unsupervised digging. supervised use means the software is directed to only look at certain files that are the most promissing candidates.