Spammers (Re: Can it help me to get a job?)
From: TLOlczyk (olczyk2002_at_yahoo.com)
Date: 12/29/03
- Next message: pete: "Re: Reverse search in a singly-linked list"
- Previous message: Randy Howard: "Re: Runs in the background, how?"
- In reply to: Richard Heathfield: "Re: Can it help me to get a job?"
- Next in thread: Bill Godfrey: "Re: Spammers (Re: Can it help me to get a job?)"
- Reply: Bill Godfrey: "Re: Spammers (Re: Can it help me to get a job?)"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Mon, 29 Dec 2003 05:39:50 -0600
On Sun, 28 Dec 2003 07:48:50 +0000 (UTC), Richard Heathfield
<dontmail@address.co.uk.invalid> wrote:
>ObProgramming: It may, alas, be possible for the effectiveness of Bayesian
>filtering to be reduced if the weight of spam far exceeds that of ham. Does
>anyone have any concrete experience of this? Is it a genuine possibility?
>And, if so, is there a fix?
>
It is a possibility. Especially if the spammers poison the filters by
putting lots and lots of good words, and by making bad words
harder to identify.
>(Because if there is, I'd like to know about it /before/ I implement BF,
>rather than afterwards!)
As a filtering method, Bayesian filtering will act very well for
anyone for a while. I also advocate blacklists ( if any still exist ).
However BF will not stop spammers from trying to get mail past
your filters and into your mailbox, and thus they will fail to be
anything but a short term palliative.
Paul Graham makes a very stupid argument that BF will make spam
to expensive for spammers continue on. But his argument has a very
major hole. His claim is that by using BF, the spammer has to send
more and more spam to get past it, and still maintain his hit rate.
Fact is that you are not decreasing a spammers rate by using BFs.
Anyone who uses BFs in the first place isn't going to buy from a
spammer, so that spam is part of the dross.
Let me illustrate this by a simple example. Let us say that a spammer
has to send 100 emails a day to make one sale. Let us say that you
get three pieces of email from said spammer each day. You decide
that you will use BF. So how many spams does the spammer have send to
make one sale per day? If he sends out a 100 a day, only 97 a day get
read.
But I will gaurantee that the spammer will still make one sale per
day. Why? Because you weren't going to buy from him anyway.
So it was one of the other 97 spams sent that resulted in his making
a sale. Of course anyone who uses BFs isn't going to buy. So if
BFs filter 99 of the spams, you still know he will make a sale,
because it was the other person who would have bought in the first
place.
Of course this is also an argument for why spammers shouldn't try
to get past filters in the first place.
--------------
The real way to deal with spammers is to strike at their anonymity.
There is a great deal of hatred directed at spammers. To the point
where acknowledged spammers have received death threats. So spammers
rely techniques to make them anonymous. In particular rooted machines
and open relays.
The way to stop spammers is to dig out these machines and bring them
down ( not by some illegal means, but by complaining to their ISPs ).
Especially the people who have rooted machines or unintentional relays
( created, for example, by installing AnalogX Proxy ).
The reply-to email address is olczyk2002@yahoo.com.
This is an address I ignore.
To reply via email, remove 2002 and change yahoo to
interaccess,
**
Thaddeus L. Olczyk, PhD
There is a difference between
*thinking* you know something,
and *knowing* you know something.
- Next message: pete: "Re: Reverse search in a singly-linked list"
- Previous message: Randy Howard: "Re: Runs in the background, how?"
- In reply to: Richard Heathfield: "Re: Can it help me to get a job?"
- Next in thread: Bill Godfrey: "Re: Spammers (Re: Can it help me to get a job?)"
- Reply: Bill Godfrey: "Re: Spammers (Re: Can it help me to get a job?)"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|