[Spambayes] Can SpamBayes be improved with Markovian Weighting or Chained Karnaugh Mapping?

[Tim Peters]

[DUI-DWI]
> More of a technical question here.  I've been doing ton's of research on
> spam in general, and am curious about the following. ...

I expect you've barely scratched the surface.  There's a huge
technical literature on algorithmic text classification, going back
more than four decades.  Spam identification is one instance of that
problem.  Dozens of schemes have been tried.  If you want to compare
two specific schemes, bottom line is that you have to try them both,
head-to-head, uning careful statistical testing.

Sorry, but I just don't have any quick answers or insights for you.