Issue 4458: Data mining metamodel                                                                                                                                                                                   (cwm-rtf)
Source: International Business Machines (Dr. Daniel T. Chang, dtchang@us.ibm.com)
Nature: Revision
Severity: Significant
Summary: This is an issue deferred from the CWM FTF. The Data Mining metamodel should be revised based on feedback from the JSR-73 (JDMAPI) work
Resolution: 
Revised Text: The CWM 1.0 Data Mining Model had the following hierarchial structure: 
 
The new model is as follows: 

 
  

 
  

 
In the above model that outlines the package dependency, notice that the various types of Data Mining algorithms have been seperated into individual packages. This new package structure is as follows: 
  

DataMiningCore -- Common classes to all Data Mining algorithms 
AssociationRules -- A type of algorithm 
Clustering  -- A type of algorithm 
Supervised -- A type of algorithm 
AttributeImportance -- A type of algorithm 
Classification -- A type of algorithm 
Approximation -- A type of algorithm
This re-factoring of the model will allow users who implement only a sub-set of the Data Mining Algorithms to participate as CWM compliant. 
The overall resolution to this issue was to replace the entire Data Mining Chapter (chapter 12) in the formal specification (document formal/2001-10-01) to reflect new metamodel definitions.


Actions taken:
August 3, 2001: received issue
May 13, 2002: closed issue
Discussion: 
Resolution: 

The CWM.Analysis.DataMining package was re-factored to allow for varied implementaions. In DataMining the various products all support some sub-set of the existing DataMining features and algorithms. The current model is fine for interchange where a user who doesn't support a specific feature or algorithm can simply ignore it on import. The CWM model forms the basis for JDMAPI which is an API specification. The structure of the model must now reflect a way to specify various compliance points based on implementation options. In order to achieve this, the DataMining model was split into finer grained packages which are based on the different DataMining algorithms. We felt this was the correct path to take, and felt justified based on the fact that to our knowledge there are no implementers of the DataMining portion of CWM. 
The resolution was to replace the existing CWM.Analysis.DataMining package with the revised one, this was distributed for review on 8/23/01 and presented in Toronto on 9/11/01.

Revised Text: 

End of Annotations:=====
From: webmaster@omg.org
Message-Id: <200108031848.f73ImHt15037@emerald.omg.org>
Date: 03 Aug 2001 14:51:51 -0400
To: <juergen@omg.org>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Subject: Issue/Bug Report
Content-Type: Text/html; charset=windows-1252
X-UIDL: j:L!!D;j!!)a1!!D9od9



Name: Dan Chang 
Company: IBM 
mailFrom: dtchang@us.ibm.com 
Notification: No 
Specification: CWM Specification 
Section: 15 
FormalNumber: ad/01-02-01 
Version: 1.0 
RevisionDate: 2 Feb 2001 
Page: 15-369 
Nature: Revision 
Severity: Significant 
HTTP User Agent: Mozilla/4.61 [en] (WinNT; U) 


Description

This is an issue deferred from the CWM FTF. The Data Mining metamodel
should be revised based on feedback from the JSR-73 (JDMAPI)
work. (This issue was already submitted. Please correct the nature and
severity of the issue per this submission. Thanks.)