Trends in Data Mining and Knowledge Discovery - Advanced Techniques in Knowledge Discovery and Data Mining - page 6

preparation, modeling, evaluation, and deployment. They called the entire process
a data mining process, which is different from the term's understanding in the
United States.
Another six-step DMKD process model [18], described below, is based on the
CRISP-DM model but differs from it in the following:
x the entire process is called the DMKD process, thus resolving the confusing
use of the term DM (it is just a step, not the process).
x the DM step is used instead of the modeling step. The DM step is concerned
with modeling and actual extraction of knowledge from data.
x several new feedback mechanisms were introduced. The CRISP-DM model
has only three major feedback sources but our experience shows that the new
feedback mechanisms are as important as the three [18], [44], [47], [61]; see
Fig. 1.3.
x the model is able to communicate with other domains; the knowledge
discovered for a domain may be applied to other domains.
Fig. 1.3. The six-step DMKD process model.