|
|
|
Overview
|
 |
Email Systems provides two different methods of defining spam: implicit and explicit.
When a particular spam filter module is enabled, this is an implicit definition, e.g. enabling Blacklist-based filtering adds .all mail originating from domains in blacklist database xyz is to be considered spam. to one.s overall spam definition. Thus administrators or users implicitly define their idea of what spam is by selecting which filters they wish to enable. Since Email Systems provide a wide range of different filtering options this allows a very flexible definition of spam.
Explicit spam filtering is implemented by means of .rules.. E.g., a rule that says .quarantine all mail that contains the word .HGH. in the subject. is an explicit declaration of a spam definition. An arbitrary number of rules can be created; rules are hierarchically arranged and each layer inherits the rules (with certain modifications allowed) from the layer above. This allows administrators and users considerable flexibility in the way that spam is defined.
Process
Currently five levels of implicit spam filtering are available:
Level 1
Every e-mail that passes through any EMF server globally is passed through a module which creates a unique signature value of each e-mail. These values are locally checked against an internal database, if a match occurs then the e-mail is classed as spam. If the value does not trigger then it is sent to a Central Spam Monitoring System (CSMS) that is located in one of EMF System.s data centres. The CSMS analyses data it receives and looks for similar e-mails being sent in bulk over a short time frame. When such an email is spotted further tests are performed and based on this the system decides if the email is spam or a legitimate mailer. If it is spam then the value is propagated out to EMF servers globally for the local spam database.
Level 2
This spam filtering method is Collective Spam Reporting (CSR). E-mails that multiple end users report as spam are broken down to a signature value and as in Level 1 the values are propagated to each EMF server worldwide.
Level 3
Makes use of a global black list database of know open relay servers.
Level 4
A Bayesian filtering module that statistically decides if an e-mail is spam based on the word content.
Level 5
A global back list database of IP addresses that known spammers operate from.
Level 6
Checks to see if the senders e-mail address is valid, if it isn.t then the sent e-mail is tagged as spam.
Reporting
Detailed reporting is available showing how much spam has been received, to whom and who from.
|
|
| Back to the Top |
|
| Overview :: Features & Benefits :: Associated Products |
|
|
Features & Benefits
|
|
Benefits
. 100% managed detection means all spam definitions updated automatically
. Multiple detection methods
. Best of breed policy means best spam protection always available
. Stop spam before they reach local networks
. Saves user time deleting spam
. False positive peace of mind with online quarantine
. 24 hour support
|
|
| Back to the Top |
|
| Overview :: Features & Benefits :: Associated Products |
|
|
Associated Products
|
|
|
|
|
| Back to the Top |
|
| Overview :: Features & Benefits :: Associated Products |
|
| << Back to product list |
|
|
|
|
