Stopping Spam This article by William Gurley mentions the problem with false positives in blocking spam. Surely, this is a big problem. However, the false positive could be solved by software that is accessible to, and configured by, the email end user. How so?  First, the rejected messages could be put into one or more spam folders, so that they could be reviewed by the users. This means that false positives need not be a serious problem, because the false positives could be reviewed by a human, perhaps on a weekly basis. Second, the rejected messages could be ranked, and automatically put into spam folders based on their rank. The idea is to reduce the amount of time that a user must expend in reviewing rejected messages. Those messages that are considered the closest to false positives could be reviewed first and more frequently than those that are strongly considered to be genuine spam. Third, there could be some kind of peer review. If any member of an organization -- a corporation, for example -- indicates that a particular message is spam, that message could be placed into the spam folder of every other member of the organization. Fourth (and so obvious, I almost forgot to include it in this list), the end user is more capable of creating effective white lists and black lists. A white list overrides the filter to not reject messages based on some criteria like the sender's address. A black list works in the opposite way, overriding the decision of a filter not to reject a message.

1:15:58 PM