The 1/3 Rule of Life (and Spam)
Initially, OT but quickly veering back to spam talk. For the sake of discussion, I've chosen 1/3 for convenience. Higher, lower or anywhere else the truth is the same.Say Mr. Greenspan and his Fed pals raise rates. 1/3 of the people panic, “OH MY GOD, INFLATION”, 1/3 are happy and 1/3 will not care.
Now say Mr. Greenspan and the Fed lower rates, 1/3 of the people panic “OH MY GOD, A SLOWING ECONOMY”, 1/3 are happy and 1/3 will not care.
Time for the obvious: You can’t make everyone happy.
Now on to spam filtering observations.
As mentioned often here, my company, CMS, has an email filter product, Praetor. One of its settings controls “spamicity”, as used by Bayesian filtering. The default value is “0.60” (scale from 0.0 to 1.00). Easily changed by an admin, an improper setting can invoke my “1/3” rule.
Set spamicity too high and unwanted mail in user inboxes increases and false positives decrease. 1/3 of the people panic, “OH MY GOD. I’M DROWNING IN SPAM”, 1/3 are happy and 1/3 will not care.
Set spamicity too low and unwanted mail in user inboxes decreases but false positives could increase. 1/3 of the people panic, “OH MY GOD. A QUARANTINED MESSAGE”, 1/3 are happy and 1/3 will not care.
(Note: Praetor’s Personal Log Viewer, easily empowers an individual to scan their quarantined messages and handle them accordingly.)
The moral: Periodically monitor your corporate email and adjust email filter settings.
Only an administrator, with knowledge of a company's internal policies and operations, can really determine this while "farming-out" these decisions could lead to problems. Admins must be in complete control. These email filter decisions do directly affect their business.
Incidently, a truism of my life is "leave as many things as possible at their default". Boy, does this reduce frustration and leave time for the really important things. (Hint, hint: Praetor at spamicity 0.60 should be just fine.)
AAS
1 Comments:
This invokes another old saw: "The squeaky wheel gets the oil."
The 1/3 that panics is going to make a lot more noise than the 1/3 who are happy, so finding the point at which the most are happy can be tricky...
Post a Comment
<< Home