The Power of a Global Database

The fact that Death2Spam has a globally-merged database really shuts the spammers down quickly. And it does that as it reduces administration time for users.

As an example, let’s say a spammer carefully constructs a new spam which gets through 75 percent of the spam filters out there, because they use simplistic first or second-generation spam-detection technology.

So this clever, nasty little spam message gets sent to the 50 million email addresses he’s just purchased on a CD from a net trawler.

Let’s say 1 million recipients of that spam are using a D2S filtering service. Within moments, 10 or 20 of those users have retrieved their mail and found the message tagged with [spam], [spam?] or [???].

If it’s [spam], D2S obviously sees this spammer’s signature profile somewhere in the message, or the website it points to.

If it’s [spam?] or [???], D2S is a little unsure at this stage. So because those 10 or 20 D2S clients don’t want to see any more of these spams from this organization, they fairly quickly open their D2S account and reclassify the unsure email as spam. The probability curve goes way up after only a few [spam] flags.

So within a few minutes of the spammer sending out his spam, the other 999,970 D2S users will not get his message. And by the fact that none of them re-classify it out of the D2S rating of [spam] in their individual databases, the probability rating gets higher and higher, so it will always be classed as spam. That spammer is globally known at that point, and any further Internet pollution that he propagates will be cut off at the knees by D2S filters across the globe.

Because of the incredibly low administration overhead of this system by the D2S users, it becomes more and more widely used as a technology. The larger the pool of D2S users, the more accurate D2S becomes due to the global database, and the more people in the group benefit by being happily unaware that the spam even existed.

All of that means the spammers’ results get poorer and poorer until they think up a new way round – and they will. But within minutes of a new spam being sent, its effectiveness is totally hobbled, due to the self-learning engine and the global database.

So, this really is death to spam.  :-)