Talk:Data mining/Archive 3

=2008=

Jane16
Today's anon edit added a link to Jane16. The link claims to be a definitive text analyzer, but contains "only english language is understand by this system" and other nonsense. - Dan Dank55 (talk) 17:16, 22 February 2008 (UTC)

Oops, wrong edit summary on my reversion
I reverted an undiscussed external link to a specific bookseller by new anonymous user 64.201.248.10, and I left a msg on his talk page inviting him to discuss the link and read WP:EL. - Dan Dank55 (talk) 04:40, 2 March 2008 (UTC)

Distinction between KDD and DM
This article seems to confuse Knowledge Discovery in Databases with Datamining. Adriaans's quote given in the introduction is actually his definition of KDD, "the nontrivial extraction of implicit, previously unknown, and potentially useful information from data". A better definition of DM is given by Fayyad: "a part of the KDD process relating to methods for extracting patterns from data".

KDD encompasses data selection, pre-processing, transformation, data mining, and Interpretation. DM is methodology used to uncover previously unknown pattern. The distinction is important as the process of KDD is not the same as the step of data mining. While I understand that in colloquial usage DM and KDD are used interchangeably, I think it is important that a factually accurate encyclopedic article makes the distinction. I therefore propose splitting this article into an article on DM and some of the algorithms ect, and an article on KDD describing the entire process. What do others think? Dmmd123 (talk) 23:49, 12 July 2008 (UTC)

Proprietary algorithms as examples
All the algorithm listed have only Microsoft algorithms as example. Why is it so? Why not list standard algorithms which are listed in open and are not associated to/created by certain company. —Preceding unsigned comment added by 203.123.175.15 (talk) 21:48, 27 November 2008 (UTC)

Naming possibly patented algorithms from a specific vendor as examples in the section "Algorithms" is questionable, specifically as now there is no added information, only a commercial. Either there has to be a solid motivation for presenting these distinct algorithms by name, or they should be removed. Hakeliha (talk) 08:57, 18 November 2008 (UTC)

Does this sound right?
The second sentence of the article says.. It is usually being used by huge corporations employing Business Intelligence essentially..

Somehow I find it doesn't sound right. --122.168.253.216 (talk) 13:39, 18 November 2008 (UTC)
 * I reworded it a little. Should look better now. Feel free to make such changes yourself in future. Anybody can edit Wikipedia. --GraemeL (talk) 13:48, 18 November 2008 (UTC)

Merger of Subject-based data mining
The topic of data mining for specific purposes is already covered in Data mining. Additing the minimal additional information in Subject-based data mining would prevent interested readers from having to read separate and overlapping articles. Bongo matic  06:02, 6 December 2008 (UTC)


 * I do not think that Subject-based data mining is of much value, its basically one quote. It seems to overlap more of with Association rule learning, which is a method of data mining. I suggest it is deleted and redirected to Association rule learning.210.1.214.216 (talk) 11:36, 19 December 2008 (UTC)


 * Both Subject-based data mining and Pattern mining are introduced in Data mining and don't seem to appear significantly anywhere else. Hence, I have merged BOTH into that section. Pdfpdf (talk) 10:44, 13 January 2009 (UTC)