Practical Data Mining

Some live blogging from HUMlab. The opening presentation for the Workshop Around Data Mining is now going on. Lars Sköld is presently giving a presentation entitled Practical Data Mining. It is centered around the program Clementine 9.0.
The definition given to data mining by Sköld was in part: “Data mining is the process of discovering new correlations, pattern and trends by sifting through large amounts of data stored in repositories”.
The base method referenced by Sköld is the Cross Industry Process for Data Mining (CRISP-DM).
Clementine seems to be a quantitative based data mining management program that uses layers of information and process which can intersect. One of the interesting points that I have picked up from the demonsration is that it is visually intensive with a icon field being the main work area. This has interesting implications to the nature of data mining technologies and can be seen as the necessity for a visual object form of representation for large bodies of information.
Initally based around quatative based methods, there is also the possiblity to undertake Text Mining that seems to work on terms and tags rather than numbers. To include tags would open up the possibilities for information processing I believe