Can a Bayesian Spam Filter Play Chess?

martin-boundary writes "The typical Bayesian spam filters learn to distinguish ham from spam just by reading thousands of emails, but is this all they can do? This essay shows step by step how to teach a Bayesian filter to play chess against a human, on Linux, with XBoard."

Comments

Re:No, it can't (well)

[panda]

Ah ha! But what if you do to the chess-bayes what you do with your spam-bayes?

Bayesian spam filters are not particularly smart at telling the difference between spam and ham until trained. That's why the directions for spamassassin and most other bayesian filters recommend that you have it learn from some saved spam and ham messages before putting the filter into production use. When that is done, the filter starts out being somewhat trained and does a much better job from the start.

I'll wager that if you ran through a set of games, say from a collection of books on championship chess, then the bayesian filter would eventually learn to play like a grandmaster. At least as far as the traditional chess canon is concerned. Make random moves against it, and it will be totally confused.