From: Baron Fujimoto (no email)
Date: Tue Sep 03 2002 - 20:49:18 EDT
On Tue, 3 Sep 2002, Greg A. Woods wrote:
: [ On Tuesday, September 3, 2002 at 09:58:13 (-1000), Clifton Royston wrote: ]
: > Subject: Re: Using blacklists and RBL's with Postfix
: >
: > Actually, Jason Rennie's ifile has been implemented and in use since
: > 1996. At first glance, it sits on the end-user delivery side, though.
:
: interesting.... and ifile works with MH et al, GNUS, procmail, Pine,
: and probably anything else with similar hooks... It's probably not a
: lot different than bogofilter except that ifile is more generic and
: tries to learn how you re-file all kinds of messages, not just spam
: vs. non-spam.
:
: > In other notes, Microsoft has been issued a patent on statistical
: > filtering of email on an application submitted in 1998, though I don't
: > have the URL handy.
: >
: > ifile seems to be prior art for everything in their claims and to
: > potentially invalidate the patent, but that would take a lawsuit to
: > establish.
:
: Though Paul Graham's publication was the first discussing this idea to
: come to _my_ full awareness, I'd bet that anyone involved with language
: analysis could have instantly thought of using Bayes' Theorem to
: identify whether some arbitrary spammer was the author of a given
: message or not. This really isn't a new idea (~1960's). I.e. there's
: likely no innovation and thus no invention for M$ to claim. (though it
: would be interesting to read their silly patent if anyone happens to
: have the number handy)
Patent # 6161130
<http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&p=1&u=/netahtml/search-bool.html&r=1&f=G&l=50&co1=AND&d=ft00&s1='6161130'.WKU.&OS=PN/6161130&RS=PN/6161130>
Technique which utilizes a probabilistic classifier to detect "junk"
e-mail by automatically updating a training and re-training the
classifier based on the updated training set
Abstract
A technique, specifically a method and apparatus that implements the
method, which through a probabilistic classifier (370) and, for a given
recipient, detects electronic mail (e-mail) messages, in an incoming
message stream, which that recipient is likely to consider "junk".
Specifically, the invention discriminates message content for that
recipient, through a probabilistic classifier (e.g., a support vector
machine) trained on prior content classifications. Through a resulting
quantitative probability measure, i.e., an output confidence level,
produced by the classifier for each message and subsequently compared
against a predefined threshold, that message is classified as either,
e.g., spam or legitimate mail, and, e.g., then stored in a corresponding
folder (223, 227) for subsequent retrieval by and display to the
recipient. Based on the probability measure, the message can alternatively
be classified into one of a number of different folders, depicted in a
pre-defined visually distinctive manner or simply discarded in its
entirety.
-
To unsubscribe, send mail to with content
(not subject): unsubscribe postfix-users
|
|
|