On Mon, 3 Mar 2003, Bob Wyman wrote:
Filters based on textual analysis cannot succeed.
I completely disagree.
If by "succeed", you mean "stop all spam", then you're right. However, I
define "succeed" as "stop a large-enough proportion of spam to save
money", and by that measure, textual analysis can be very effective.
I believe that there is no solution that will totally stop spam. What
is needed is a broad-spectrum approach consisting of:
- A legal framework to make it possible to prosecute spammers and extract
money from them.
- Technical non-content-based means to reduce spam, such as RMX records.
- Buy-in by ISPs to take technical measures against spammers (ie,
force subscribers to relay mail through the ISP's SMTP server.)
- Text-based analysis to detect likely spam.
- Tools to increase spammers' costs, such as tarpitting, tempfailing
suspect messages, etc.
Like it or not, this is an arms race. The best we can hope for is to increase
spammers' costs without hurting ourselves too much in the process.
Let's focus on
fundamental solutions to the problem of spam -- i.e. how to build
systems for "consent-based" messaging.
In my business, I rely on getting e-mail from completely unknown
people. This is a feature of e-mail I cannot do without. Anything
that makes it harder for strangers to contact me than simply dashing
off an e-mail is a non-starter for me.
--
David.
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg