|
|
Spam, junk email, bulk email, UCE are all basically the same thing.
It can be annoying, offensive or even dangerous. Like most internet users,
you would prefer not to get it. We can help. Our spam filters are constantly
being updated to ensure that we are being as effective as possible
without stopping the flow of any legitimate email.
Here is how we do it...
Your email is forwarded to our network where our system analyzes it.
We use a number of different methods to determine whether or not an
email is spam. Our system is much more effective than any system that uses only
one method of spam detection. All
email that enters runs through approximately 750 different
rules and tests (some methods are described below). A score is determined and the email is either
deleted or passes through, based on its "spam probability". This
system is proving to be very effective and you will notice a significant,
immediate reduction of junk email.
| Pattern Matching | Pattern matching is a method of
describing a message in terms of its content. The spam filter
examines layout and organization, to identify the common
characteristics of spam. An advanced pattern matching engine simultaneously
applies thousands of algorithms during a single pass. The results
determine a probability rating, and assign the appropriate action.
Patterns are updated regularly to identify new tactics. |
|---|
| Spam Definitions | Spam definitions are patterns used to
identify specific instances of spam. They detect multiple instances
of material considered sufficiently similar to be the same message;
for example, hoaxes and chain letters that may otherwise be
difficult to detect. Spam definitions are updated regularly. |
|---|
| Real Time Black Lists | The Real Time Blackhole List (RBL),
DSBL and DNSBL are compilations of networks that either allow
spammers to use their systems to send spam, or have not taken
action to prevent spammers from abusing their systems. |
|---|
| Heuristic Analysis | Heuristic analysis uses a series
of internal tests to determine the likelihood that a message is spam
. Each test is weighted with a point value to reduce false
positives. The total probability of spam is examined to determine
an overall score, and a mapping function assigns the appropriate
action. |
|---|
|
