KDD procedures – KL-Miner

 

KL-Miner procedure

Author(s):

Milan Šimůnek, Jan Rauch

Responsibility:

Jan Rauch (theory), Milan Šimůnek (software), Martin Kejkula (help)

Description:

Data mining procedure KL-Miner mines for patterns of the form R ~ C / Cond. Here R and C are categorial attributes, the attribute R has categories r1, , rK, the attribute C has categories c1, , cL. Further, Cond is a Boolean attribute [RSL 03].

The procedure deals with data matrices. The attributes R and C correspond to columns of the analysed data matrix. Boolean attribute Cond is derived from the other columns of the data matrix.

The intuitive meaning of the pattern R ~ C / Cond is that the attributes R and C are in relation given by the symbol ~ when the condition given by the Boolean attribute Cond is satisfied.

The symbol ~ is called KL-quantifier. It corresponds to a condition imposed on the contingency table of R and C. The pattern R ~ C / Cond is verified on the contingency table of R and C in data matrix M / Cond. Here M is the analysed data matrix and M / Cond is a data matrix consisting of all rows of M satisfying Cond.

There are various KL-quantifiers, some of them correspond to simple conditions concerning frequencies, and some others are of statistical nature e.g. χ2-statistic.

There is a demonstration of the main features of the procedure KL-Miner.

Files to download:
LISp-Miner.Core.OldUI.zip 33.45 MB August 13, 2014
Legacy LISp-Miner system core files separated into modules for each GUHA procedure. Contains also other legacy modules LMAdmin and LMDataSource.
History:

The procedure KL-Miner was suggested by J. Rauch in 2003. Reason was the necessity to mine for patterns describing a relation of two categorial attributes in a more concise way than it is possible by associational rules. The second reason was the possibility to use the software tools for dealing with strings of bits developed for the 4ft-Miner procedure.

The procedure was implemented by M. Šimůnek. Several KL-quantifiers were suggested by P. Berka and by V. Lín. V. Lín and P. Dolejší solved lot of tasks with the KL-Miner and suggested several improvements of the procedure [RSL 03], [RSDL 04].

Print page

 
 

Send comments about this site to the webmaster