|
Spam
filters are used by most people to avoid getting spam
into their inboxes. Most e-mail provides have free e-mail
filters included with every e-mail account.
This
is to the benefit of the customer and the provider, because
it keeps annoying spam away from the customer, and it
prevents the provider from having to give the customer
an excessive amount of room in the inbox.
Many
people prefer spam filtering to blocking, because all
suspected spam goes the bulk folder and is subject to
the user's inspection before it is deleted or accepted.
This allows the user to accept e-mails that the filter
considered spam and were not.
False positives (suspected spam that is not spam) usually
occur more often than false negatives (spam that is not
regarded as such by the filer and ends up in the inbox).
Because of occasional errors that are inevitable, many
users prefer filters as a less radical approach to keeping
spam away.
Filters
work according to Bayes' theorem, which , as it applies
to spam, deals with calculating the probability that e-mails
containing certain kinds of word or wording are spam.
These buzz words and phrases are compiled after much research
on the creation of spam, and habits of spammers are studied
before calculating these probabilities.
The
Bayes formula is highly effective in allow spam filters
to get rid of most types of spam, but many spammers are
getting smarter and are avoiding using words that have
a high probability of appearing as spam.
This
ever-evolving spamming vocabulary, is a problem for creators
of spam filters who must constantly see to it that their
product evolves to the point where it is a step ahead
of the spammers. This is not easy, since most spammers
are computer saavy, and can spend all day finding ways
to get past spam filters. In general, it is easier to
slip under a wall than to engineer a foolproof wall.
However,
spam filters, while not flawless, are still the moderate
choice for many who want to avoid receiving a large amount
of spam, but who do not want to go as far as allowing
only certain users to e-mail them. These computer users
usually receive a dozen or so spam e-mails a day, as opposed
to the hundreds of spam e-mails that invade the inboxes
of small business owners.
Spam
filters are expected to become more powerful as spammers
become more ingenuous. For instance, certain words that
have a high probability of appearing on spam e-mail, such
as Viagra, are often deliberately misspelled to something
like Viaagra. This can fool many filters, but for users
who recognize this fact, it is easy to spot this kind
of spam. The user can also train a filter to avoid e-mails
with certain words and phrases and can permanently block
messages from certain recipients.
The
problem with this is that many spammers use spam zombies
whose e-mail addresses are forced to send spam. A spammer
can be blocked only to have the same message sent through
the spammer spoofing an acquaintance's e-mail address
or from another PC.
Although
spam filters are not perfect, they are still well worth
trying out, since they are effective at filtering out
the vast majority of spam e-mails.
|
