More than 50% of users regularly double-check for false positive spam filtering


Only 12% of users trust their spam filter sufficiently not to bother sifting through spam folder.

In a poll of more than 1,000 visitors to the

VB

website, 52% of users say they regularly check their spam folder for false positives, while only 12% of users said they never bother to check for legitimate messages misclassified as spam.

The subjective nature of spam (one person’s spam may be another person’s legitimate email), along with the ever-changing nature of spam as spammers come up with new tricks and techniques to bypass filters means that spam filters are unlikely to have 100% accuracy – indeed even the term ‘accuracy’ is

considered to be flawed

when it is used in reference to spam filters.

One might wonder, then, at the wisdom of the 12% of respondents who say they never check for false positives. Another 26% of respondents said that, while they check their spam folders, they only do so occasionally. Which of these policies is best depends on the spam filter used and the way in which it is configured, as well as on the nature of the email account in question: for personal correspondence, a missed email might simply be inconvenient, whereas a company that misses out a sales lead because an email was incorrectly identified as spam may lose important business.

In either case, it shows that spam is more than just an annoyance of the digital age. Missed emails could mean missed opportunities, but with more than 90% of email traffic estimated to be spam, many employers are equally concerned about the amount of time employees are having to spend on sorting the ham from the spam.

The full poll results can be found

here

.

An article describing the challenges in spam filter evaluation can be read

here

(pdf, p.21). And an article on measuring an marketing spam filter accuracy can be read

here

.

Posted on 16 April 2008 by

Virus Bulletin


Posted

in

by

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *