Skip to Content
Author's profile photo Former Member
Former Member
July 27, 2006 Less than a 1 minute read

Spotting trouble via text mining

Text mining is perhaps a $100 million industry (maybe a little less only counting software license fees, but certainly a bunch more if one counts more liberally).  A large fraction of that business, and of its applications, can be grouped under one umbrella — spotting signs of trouble.  For example (and this is not even a complete list), text mining has been used in:      * Vehicle safety     * Other manufacturing/warranty analysis apps     * Reputation management (for the most part)     * Other customer sentiment apps (some, perhaps most)     * Anti-terrorism     * Sarbanes-Oxley compliance     * Antifraud     * Stopping money laundering     * Clinical applications (some)     * Early insurance risk management apps     * Early experimental hedge fund apps  If there’s one area to look at for introducing text mining to your enterprise, it’s this one.  At the link above, I went into a little detail about how it’s actually done.

Assigned Tags

      3 Comments
      You must be Logged on to comment or reply to a post.
      Author's profile photo stuart jameson
      stuart jameson
      What's the approach taken to text mining in SAP where you seem to have to know the DB tables where the data is stored?  Is there any product that googlifies SAP?
      Author's profile photo Karsten Hohage
      Karsten Hohage
      Please check the information on Enterprise Search that you will find here:
      https://www.sdn.sap.com/irj/sdn/developerareas/itpractices
      Author's profile photo James Chau
      James Chau
      It is not difficult at all. I have deployed ABAP OO technologies behind the select-option to handle the complexities by providing abstraction and polymorphism at Synopsys Inc's CRM transaction monitor analytics. The results are very high performance with a lot higher accuracy than the T-Rex implementation by SAP AG.