Network and Systems Services, Computing and Information Services, Texas A&M University
blue horizontal rule

NETWORKS

SYSTEMS

CIS

TAMU
blue horizontal rule
SPAM Scanning

SPAM Scanning

The SPAM scanning is based on SpamAssassin. There are many rules that are considered for deciding whether it is likely that any given message is spam or not. Unfortunately, even with the best scanners, all spam is subjective according to the recipients as to whether they consider it spam or not. We therefore will not perform any blocking for spam at this time, but will pass it on the the recipient with an added "X-Perlmx-Spam:" header listing the reasons the scanner found that are likely to make it spam.

For the rules that are used in an attempt to identify spam, check the Table of Spam Tests .

Since messages that have a greater than 30% chance of being spam will contain the added header, mail servers and clients that offer the ability to perform automatic filtering can use that header according to the recipient's trust level in the scanner. Should the scanner be trusted to correctly identify all spam with a probability of a certain percentage or higher, a rule can be constructed to automatically file those suspect messages in a spam folder, or even be automatically deleted. Please see the Example Filters to see if there are some ideas that might help automated filtering.

[ NETWORKS | SYSTEMS | CIS | TAMU ]