for a measurement experiment, i would like O(100k) *headers* from spam
from europe and a similar sample from the states.
this would be a straight sample, before filtering, ip address blocking, etc.
if you can help, please drop me a note and we can discuss how the sample
is taken and how delivered.
thanks!
randy
Request for clarification: do you mean "spam originating at IP addresses
believed to be in Europe" or "spam received at a mail server located in
Europe" or "spam putatively from domains in Europe" or something else?
---Rsk
One thing that happened when I moved to Europe and started doing
business in Germany is that relatively soon I began receiving spam in
German (which seems to have quite different content, and sales
strategy, actually, perhaps reflecting cultural differences in the
manner of buying and selling between the anglophone world and Germany).
Trying to separate out what "in" Europe means in this case seems to come
down to having given out email addresses to web sites and collegues in
a different language environment rather than physical presence of either
myself or my mailserver in either North America or Europe. I guess the
German spam I have been receiving is only european in that German
speakers happen to be mostly in Europe, which is not true of English
speakers.
I wonder, is the (English language) spam set that one is likely to receive
in Australia statistically different than what one is likely to receive in
the US?
-w
for a measurement experiment, i would like O(100k) *headers* from spam
from europe and a similar sample from the states.Request for clarification: do you mean "spam originating at IP addresses
believed to be in Europe" or "spam received at a mail server located in
Europe" or "spam putatively from domains in Europe" or something else?One thing that happened when I moved to Europe and started doing
business in Germany is that relatively soon I began receiving spam in
German (which seems to have quite different content, and sales
strategy, actually, perhaps reflecting cultural differences in the
manner of buying and selling between the anglophone world and Germany).
I receive serious amounts of spam in Hebrew and Russian, and haven't even been to
either Israel or Russia recently.
Regards
Marshall
Rich Kulawiec wrote:
for a measurement experiment, i would like O(100k) *headers* from spam
from europe and a similar sample from the states.Request for clarification: do you mean "spam originating at IP addresses
believed to be in Europe"
yes.
and, because i have gotten a lot of well-meaning but non-reading offers,
to repeat
this would be a straight sample, before filtering, ip address
blocking, etc.
i realize this is difficult, as all of us go through much effort to
reject this stuff as early as possible. but it will be a sample
unbiased by your filtering techniques.
randy
s/recently/ever/
I'd be happy if I could tell Gmail to delete anything in a non Roman
character set. I don't read Hebrew, Arabic, Kanji, Hangul, Cyrillic, or
any of the other various character sets I get spam in.
Request for clarification: do you mean "spam originating at IP addresses
believed to be in Europe"yes.
<blush> aiiii! speaking of non-reading <blush>
i mean spam arriving at port 25 on a european host. and an unfiltered
unblocked port 25, no dnsbl, ...
it looks like i have a great stateside volunteer source, though the
proof will be known when we have the data. and we're in asia and have
data from here. so it's europe i need.
randy
Randy Bush <randy@psg.com> writes:
this would be a straight sample, before filtering, ip address
blocking, etc.i realize this is difficult, as all of us go through much effort to
reject this stuff as early as possible. but it will be a sample
unbiased by your filtering techniques.
How do you classify email as spam without adding bias?
Bjørn
From: owner-nanog@merit.edu [mailto:owner-nanog@merit.edu] On Behalf
Of Marshall Eubanks
Sent: Thursday, April 10, 2008 9:39 AM
To: William Waites
Cc: Rich Kulawiec; North American Network Operators Group
Subject: Re: spam wanted
[ clip ]
I receive serious amounts of spam in Hebrew and Russian, and haven't
even been to either Israel or Russia recently.Regards
Marshall
I started getting spam in Icelandic < 24 hours after my account was set
up. I get Russian, Chinese, and Hebrew spam all the time. The most spam
I receive is from an old domain that I turned off the MX records. Every
now and then I turn them back on to see what's flowing and it never
changes. Within seconds.
[obOp] I think that the language change defeats many of the heuristics
found in common spam appliances.
this would be a straight sample, before filtering, ip address
blocking, etc.i realize this is difficult, as all of us go through much effort to
reject this stuff as early as possible. but it will be a sample
unbiased by your filtering techniques.How do you classify email as spam without adding bias?
reasonable question. i suspect you pull out the 0.5% of the inbound you
actually wanted and consider the bias small. as the dnsbls alone block
way over 90% of the inbound here, i would not classify that as small.
randy
