Re: Best way to classify...

Bill Yerazunis <[email protected]>
Newsgroups gmane.mail.spam.crm114
Message-ID <[email protected]>
   Say, for example, that I have this piece of text which might belong to
   one of four classifications.=20

   Maybe it is spam, normal ham, ham about cats, ham about dogs. Can
   crm114 choose from multiple classifications at once?

   Or is it a binary decision sort of thing where I have to build a
   binary decision tree and consider spam vs ham and of the ham consider
   ham about cats vs ham about dogs?

   Is one more accurate than the other?

It Depends.

In general, I'd expect that doing a four-way classify would
give you the best results, but that may or may not
be true (you may be better off splitting in a different order).

I've been working on the optimal theory for this, and the 
theoretical results based on hyperspace are that no generalization
can be made; it all really depends on your source(s).

So- test it.  But my money is on a four-way classify.

    - Bill Yerazunis

------------------------------------------------------------------------------
This SF.net email is sponsored by:
High Quality Requirements in a Collaborative Environment.
Download a free trial of Rational Requirements Composer Now!
http://p.sf.net/sfu/www-ibm-com
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.