Proposals for Incorporating Machine Learning in Mozilla Firefox

Friday June 18th, 2004

Blake Ross writes: "I will be doing research this summer at Stanford with Professor Andrew Ng about how we can incorporate machine learning into Firefox. We're looking for ideas that will make Firefox 2.0 blow every other browser out of the water. People who come up with the best 3-5 ideas win Gmail accounts, and if we implement your idea you'll be acknowledged in both our paper and in Firefox credits. Your idea will also be appreciated by the millions of people who use Firefox :-). We'll also entertain Thunderbird proposals."

#68 Re: Thunderbird automatic filtering

by phaasz <>

Tuesday June 22nd, 2004 7:25 AM

You are replying to this message

Creating categories automatically would probably be quite unintuitive because they are so subjective. I certainly concur that categorisation is much more useful than foldering. Perhaps a learning Bayesian filter that learns categories instead of spamminess (as currently implemented) or folder (as suggested by gwalla above)? The user sets up whatever categories make sense (with some default categories, one of which would be "Spam").

The engine would automatically assign categories to each piece of incoming email. Marking a message as spam would function as it does now. Similarly marking a message as "work related", or "funny", or "legal action" or "job applications" (or some combination) would work. For every category, every message either belongs or does not belong. By having the user set all applicable categories with incoming email, the engine would be trained.

I would envisage that as it goes, the engine would guess all of the categories for each incoming mail article, and that this categorisation could be corrected as needed by the user to further teach the filter.

The notion of folders could then disappear. Instead, one would simply configure arbtrary "views" which are effectively stored searches (with category, date, sender, etc available as filters), groupings/sorts (conversation, sender, date, size, etc).

The junk mail folder is simply a view that shows all messages with the junk mail category set...

You'd probably want to attach a pretty mean indexing engine so that it could be searched effectively.

I hope all of this makes sense!