> Specifically, what the numbers mean. I am seeing .9xxx and 1.000 in the > header. Does this mean 97% chance it is spam and 1.00 is a 100% chance it > is spam?
> X-IMAIL-SPAM-STATISTICS: 0.9787
You are correct.
Actually (Ipswitch can correct me if I am wrong), this is not a true probability but a "pseudo-probability" (a number between 0 and 1 that acts like a probability, but isn't one). With naive Bayes theorem, for example, you will often get E-mails that show a 99.9999% chance of being spam (that's 1 in 1 million chance), whereas the real probability might really be something like 99.8% chance (1 in 500 chance). Naive Bayes Theorem essentially simulates extremely complex calculations (too complex to be done in real-time), and ends up coming up with estimates that are normally very close numerically to the actual probabilities.
In some situation, exact probabilities are much better than estimates. For example, with web hosting, the difference between 99.5% uptime and 99.99% uptime is drastic. However, for spam control, the difference between 99.5% and 99.99% isn't very important (since many people would delete the E-mail in either case). In this case, estimates are much more acceptable.
The key here is not to run any true statistical calculations based on the number you see (for example, don't assume that if you see 1.0000, that you will have a 0% false positive ratio, or that with .9990 you'll have 30 false positives a month if you process 1,000 E-mails/day).
-Scott
---
Declude JunkMail: The advanced anti-spam solution for IMail mailservers.
Declude Virus: Catches known viruses and is the leader in mailserver vulnerability detection.
Find out what you have been missing: Ask for a free 30-day evaluation.
--- [This E-mail was scanned for viruses by Declude Virus (http://www.declude.com)]
To Unsubscribe: http://www.ipswitch.com/support/mailing-lists.html List Archive: http://www.mail-archive.com/imail_forum%40list.ipswitch.com/ Knowledge Base/FAQ: http://www.ipswitch.com/support/IMail/
