Trends in Data Mining and Knowledge Discovery - Advanced Techniques in Knowledge Discovery and Data Mining - page 19

Fig. 1.6. DM toolbox architecture using the Internet via HTTP, XML, SOAP, or
XML-RPC and UDDI.
The business community already tries to integrate the DMKD process. During
the last few years businesses showed growing interest in DMKD. The biggest
DBMS vendors like IBM, Microsoft, and Oracle integrated some of the DM tools
into their commercial systems. Their goal is to make use of DM methods easier, in
particular for users that use their DBMS products. IBM's DM tool, called
Intelligent Miner, which integrates with DB2, consists of three components:
Intelligent Miner for Data, Intelligent Miner for Text, and Intelligent Miner
Scoring [58], [59]. Intelligent Miner for Data uses clustering based on Kohonen
neural network, factor analysis, linear and polynomial regression, and decision
trees to find associations and patterns in data [17], [28], [43]. Intelligent Miner for
Text includes a search engine, Web access tools, and text analysis tools. Intelligent
Miner Scoring is the DM component designed to work in real time. Intelligent
Miner incorporates some data preprocessing methods like feature selection,
sampling, aggregation, filtering, cleansing, and data transformations like principal
component analysis [17]. It also supports the PMML format. Microsoft's
SQLServer2000 incorporates two DM algorithms: decision trees and clustering
[69]. The implementation is based on the OLE DB-DM specification. Oracle has a