Bill Myers

I get a lot of email. In fact, since January I have received 16,115 emails. 39% were work related, 59% unsolicited ads, and 9% solicitations for porn.

I know the exact percentages because I use a 'smart' email program to review my email and sort it into different mail boxes for review or deletion.

The program uses 'bayseian logic' to classify the messages - a technique involving comparing words and phrases used in an email, to words and phrases in messages I have identifed as either being work related, spam or porn.

This 'smart' filtering program actually 'learns' - and it currently is showing a 99.18% accuracy rate in correctly classifying my email. (Not bad for a free program!)

On the few rare occasions that a message is misclassified, it is almost always a work related message that has been classified as spam.

Funny thing . . . almost all the mis-classified email comes from the same few people. It appears that their style of writing and vocabulary is so similar to spam messages that even a smart filtering program can't tell the difference.

For those people, this is a big problem. It means many of the messages they send will be deleted by spam filters. And it means that even if their messages do get read, they often won't be taken seriously - too close to spam.

If you don't use a filter program to sort your email, it might be a good time to start thinking about using one. Not only will you reduce the amount of spam you have to deal with, you might learn something about how to write email messages and web pages so they aren't confused as being spam.