Google Message Security for Google Apps Administration Guide - Manual (Page 169)

Chapter 9
Spam Filters 169
Spam Filters
Chapter 9
About Spam Filters
Your message security service detects spam by applying hundreds of rules to
each message that passes through the data center. It can block obvious spam
immediately, then divert more borderline spam to a Quarantine for later
evaluation. From there, you or your users can review the Quarantine for any
legitimate messages that were falsely quarantined and need to be forwarded to
the user’s Inbox. Otherwise, spam is deleted automatically.
When your service is activated, all types of spam are typically filtered at a uniform
level of aggressiveness. One group of users, however, might have its own idea
about what constitutes spam, or how aggressively to filter it. A travel agency might
have a zero-tolerance policy for adult content, for example, but want to receive
special offers, such as “trips to Hawaii.” Another group might want to change its
spam disposition, by changing how its spam is quarantined, or not quarantining it
at all.
Filtering aggressiveness affects how the protection service handles messages
that may or may not be spam. More aggressive spam filter levels will quarantine
messages that are borderline cases. This will cause more spam to be caught, but
may increase false positives. More lenient spam filters will allow borderline
messages through, which reduces false positives but potentially lets more spam
through.
For each of your organizations, you can adjust the overall aggressiveness of
filtering, filter specific categories of spam more aggressively, and choose a spam
disposition. Some of these settings are made at the org level, and some for a
Default User. You can also adjust individual user’s filtering, or allow users to do
this themselves at the Message Center.