KDnuggets : Polls : 2000 : Data Mining Software (May 2000)
Do you use [400 votes total]
Angoss (27) 7%
CART (27) 7%
Clementine (79) 20%
IBM Iminer (37) 9%
Megaputer (5) 1%
SAS (71) 18%
SAS EM (55) 14%
SGI Mineset (19) 5%
Other (80) 20%


Other tools mentioned in the other category included
  • Alice Isoft (part of Business Objects)
  • DataEngine
  • GainSmarts
  • MARS
  • MATLAB PLS toolbox
  • Model 1
  • Oracle Darwin
  • NeuralWare's Professional II+
  • SPSS 10.0 and AnswerTree
  • WhiteCross

Tool Comments, or Where is My Tool?

  • SAS
    • SAS Enterprise Miner has a great graphical interface for iterative mining, allowing business analysts to utilize powerful modeling techniques with good default parms, but also giving the PhD expert-types the control for the advanced stuff.
    • SAS EM is fast and has a great collection of algorithms - but the user interface needs much polishing.
  • SPSS
    • Where is AnswerTree and SPSS 10.0? I use these for my data Mining activities. Answertree has CHAID, C&RT, and Quest for pretty visual decision Trees.
    • Clementine is relatively easy to use and provides good breadth of capabilities at a reasonable price. Everything about SAS is too expensive. IBM has a lot of capabilities, but it's harder to effectively use.
    • Clementine is very good and has tremendous potential. The company was recently purchased by SPSS, giving it increased momentum and exposure.
  • Other
    • Where is Alice from Isoft (http://www.alice-soft.com)? Alice is imbedded in Business Objects thereby totalling more users than any other package.
    • DataEngine offers many Data Mining techniques!
    • why is GainSmarts not on it? It won the KDD Cup twice and 45 companies are using it in USA, Europe and Asia.
    • IBM IM is best, it's collection of function is richest and it's IBM
    • Use Search in Microsiris

General Comments

  • finally a poll I will consider participating in, since the sampled population (KDD types) are actually the ones who have some specific knowledge of the area.
  • It would be useful if a survey such as this would discriminate types of user, eg data mining technician, market analyst, ...
  • Easy to use, good interface but algorithms need to be more sophisticated, e.g. pruning of trees. So need better visualisation of learned models

KDnuggets : Polls : 2000 : Data Mining Software (May 30-June 12, 2000)

Copyright © 2002 KDnuggets. Subscribe to KDnuggets News!