A
"Bayesian filter" is a scientific form of spam content blocking
A
Bayesian filter uses a branch of science, known as statistics, to help
identify spam. This software filter creates two kinds of tables: one is
for words that routinely appear in legitimate e-mail, and the other for
words that appear in spam, and it assigns each word a "score."
As a "for instance," the word "Viagra" is given a score that
indicates
that it's more likely to appear in a junk e-mail spam message, as
opposed to a good
word like "meeting" which could easily appear in a good e-mail message.
Then, after scanning the entire message, a "score," i.e., a "spam
score" is assigned. Over time, as more and more e-mail is scanned, the
Bayesian filter "learns" and can become a really good anti spam filter.
Now, keep in mind that spammers are smart, and they are well
aware of
Bayesian filters. Some spam messages may contain a paragraph or more of
text that may sound quite legitimate. Spammers may include such text in
an effort to defeat the Bayesian filter by reducing the overall spam
score.
If you have anti spam software that uses a Bayesian filter
you must
take an active role in making it as effective as it can be. You must
"teach" it about spam that "leaks" thru the filter. This may require
you to "point and click" so as to add new words to the spam table.
AntiSpam Outlook | Filter Outlook | Outlook 2010 Spam Settings | Outlook 2010 Junk Email Filter false positive | Block Email Outlook 2010 |