No one can deny that junk email, or "spam", is an annoyance that's on the rise. But where many corporate email services simply censor all messages that look like spam, it's not such a simple matter in an academic community, where we assume that individual members will wish to make their own decisions about what information they keep and what they discard.

To date, Computing and Network Services has dealt with spam mainly by using the Distributed Checksum Clearinghouse software (DCC) to refuse email from well-known spam sites. DCC works in a manner similar to anti-virus filters, checking incoming email messages against a database of known spam messages and deleting the incoming message if it matches. Since the DCC database is being constantly updated from thousands of sites, and a message has to be reported hundreds of thousands of times before it is considered to be spam, be assured that only spam messages are deleted by DCC. CNS also blocks selected addresses when reports of abuse are received.

Since spammers can also easily use unprotected mail systems at legitimate sites to relay their bulk mailings, CNS has taken the additional step of blocking mail from any open mail relay that has been used to propagate spam. An organization called MAPS (Mail Abuse Prevention System, http://mail-abuse.org) maintains a minute-by-minute database of open mail relay systems being victimized by spammers and assists system administrators to fix the problem.

But despite these efforts, spam still accumulates in our mailboxes. So how to reduce the tidal wave of junk mail? Answer: use SpamAssassin, which allows you to define what is "junk" and what isn't.

Emailstats Show Spam Being Identified

DCC and SpamAssassin are catching junk email. Need proof? Take a look at the statistics online at http://emailstats.yorku.ca - they tell an interesting story. Started late in January 2004, emailstats shows the total number of email messages processed at York each day - a staggering amount - and the number of messages which are identified daily by DCC and SpamAssassin together.

Take March 11, 2004 as a typical day:

Messages processed: 603,238
Rejected email: 51,790
DCC identified spam: 47,428
SpamAssassin identified spam: 40,522

This means that on the typical day we chose, DCC and SpamAssassin together identified nearly 88,000 spam messages.

Join York's SpamAssassin users

Again on March 11, 2004, of 47,785 mail users nearly 2,000 were using SpamAssassin, which identified just over 40,000 messages as spam. SpamAssassin usage at the University has increased from 1,726 users on 22 January to 1,950 by March 24 of this year.

If you'd like to join the growing ranks of SpamAssassin users, go to the CNS website at http://www.cns.yorku.ca/computing/email/spamassassin.html for instructions on configuring SpamAssassin to work for you.

<< newsletter home