spam statistics

[Mr. W.Y. Chan]

Cassandra Smith (clsmith at DARWIN.BU.EDU) wrote:
: Any suggestions about a word other than women? What is the latin
: or Greek for women?...

: Cassandra


:    From: BIOSCI Administrator <biosci-help at net.bio.net>
:    Reply-To: biosci-help at net.bio.net
:    Cc: biohelp at net.bio.net, biohelp at net.bio.net

:    Cassandra,

:    I thought this might interest you and your newsgroup.

:    As part of an analysis of how best to configure our anti-spam efforts,
:    I did a rough analysis of the spam hitting our newsgroups. I took May
:    '97 data and looked at a sample of 5 newsgroups, large and small. 

:    Newsgroup    all_mail spam    % spam

:    METHODS	        901     29         3
:    BIOPHYSICS      155     28        18
:    PLANTS-EDUC     122     34        28
:    WOMENINBIO      117     41        35
:    ZEBRAFISH        61     30        49

:    Totals:        1356     162	

:    Average number of groups a spam message sent to: 5.00

:    Broken down by:
:    -- 31 multi-occurrence sets account for 155 spam messages			
:    -- 7 sets of 1 occurrence  (all in WOMENINBIO)		

:    To me the surprise was that the women-in-bio newsgroup is getting more
:    spam than other groups and spam targeted, presumably because 'women'
:    is in the newsgroup name, to women.


That is not entirely correct. If you take into other newsgroups other
than the ones you had already mentioned this is nothing, I have seen
ng were there had been over 5,000 posts in a month and 80% of it is spam,
the "women in bio" ng is one of the smaller ng around so any amount of
spam that get sent here would obviosely boost the percentage level over
other bigger ng. A very simple mathematical logic....:)


	Wah


:    As you can see, roughly speaking, all groups are getting about the
:    same number of spam messages, but for groups with a large volume of
:    mail it gets lost in the noise, whereas the smaller groups are
:    overwhelmed with junk.

:    Serge