Using BF for scoring text on other types of polarities?

[RW]

On Wed, 24 Jun 2009 15:42:16 -0400
jkinz at kinz.org wrote:

> Hello.
> I know people in this list have spoken about using BF for other
> purposes besides detecting SPAM. 
> 
> Are there any tools or documents, emails etc.. that give any
> hints about how you can do this? 

I don't think you have to do anything special, a text file is just like
a text email with no headers, so just follow the instuctions  for
dealing with files in maildir. 

The only issue is whether it's the best tool for the job. Like most
statistical spam filters Bogofilter just looks at whether tokens are
present, not how often they occur within the document. I think some
other kinds of textual comparison do use this information.