Types of Spam Filters

Spam filters are designed to block unwanted messages generated
by unethical senders. Presently, there are different spam filter
programs in the market available for email users to purchase.
These programs are designed differently, which means that their
methods of detecting spam vary.

Businesses which require email messaging for information
dissemination would find it helpful to know the different
methods used by spam filters. This will in turn help them tailor
there message's content as well as their mailing procedures.
Thus, being aware of spam filter programs can help in the
optimization of any internet campaign.

The most common type of spam filter is the Content-Based Filter.
This type of filter as the name suggests, looks on symbols,
words or phrases in the subject line, headers and content of the
message which are suspicious. The filtering process is very
dependent to phrases and words chosen by the user as spam, this
means that spam messages with words or phrases not included in
the list of the user will make it to the inbox time and again
unless the user creates another filter to deal with those words
and phrases. The good side of this filter is, it can be
fashioned in anyway the user wishes to.

Another type of spam filter is the Bayesian Filter, unlike
content-based filter; Bayesian filter undergoes 'training' where
the filter learns the difference between a spam and a legitimate
email. During this 'training' messages are broken into tokens
and these tokens are then stored in databases (tokens of good
messages are separated from those of spam). This type of filter
requires minimal maintenance and can adapt easily to users'
usual message, just like if the user is a medical practitioner,
anything that relates to medical field will be considered as
good. The primary drawback of Bayesian filter is that optimum
filtering will not take place immediately, since the filter is
dependent on the messages on which it is trained.

In addition, there is a Blacklist-Whitelist filter where the
whitelist is composed of the 'good' email addresses and the
blacklist is composed of email addresses believed to send spam.
This kind of filter ensures that received emails are from
whitelisted senders only and any address aside from the
whitelisted ones will be blocked. With this, email users can
expect a zero false negative. The main disadvantage of this
filter is that, whenever a real-time blacklist (RBL) is used,
there is a tendency that the whole IP is considered as spam
sender, though only a part of it used to send spam.

Challenge-Response Filter on the other hand sends messages to
unknown senders asking them to do further action/step to justify
their intention of sending the email. These further actions can
be a request for the senders' website to ensure that the email
is legitimate. This type of filter is quite applicable to low
traffic email users.

Community filter is another type of spam filter. This filter has
a central server where users can send suspected spam; the server
itself then saves the 'fingerprint' of the message to the
database. When enough users flagged the message as spam, the
'fingerprints' of the message will be blocked in the future. But
before a spam is blocked, many users have already opened it.

If your business makes use of email in sending promotional or
newsletters, you should at least know if your email address is
in the real-time blacklists so that you can apply for
whitelisting. And at least know the basic symbols, words and
phrases to be avoided in sending emails. Try visiting href="http://www.emailreach.com">www.emailreach.com for
more information about spam filters.

About the author:
Registered Electronics Engineer

Author: Roderick
Article: