Decius wrote:

noteworthy wrote:
Whose reputations? Not users'. Instead, it's server admins and email advertisers.

I haven't read this, but generally speaking the problem with using "reputation systems" as a spam filter is the new user initiation problem. If people wanted new users to face hurdles they could easily deploy whitelisting with challenges today. They aren't, either because they are satisfied with the amount of spam they get, or because they perceive that challenging new emailers costs more than spam.

Well, you have to remember that servers are owned by humans and many servers may be owned by a single human (or group of humans).

emailRM joins a human's reputation to a server's.
The servers are categorized in 2 buckets, advertisers, which include spammers and non-spammers, and non-advertisers, which include anyone that isn't advertising or mass emailing something.

If a server is unclaimed by a a human, it is still classified into one of the 2 buckets, but will have the worst possible reputation.

When someone claims ownership of a server (By the way, ownership is verified with the ISP) the information provided by the owner will determine the server's reputation, including any other servers the owner successfully claims as his.

If the server is in the advertiser bucket, the server keeps a bad reputation until the owner submits himself to a marketing assessment in which it is determined if the owner is a reputable advertisers.

If the server is in the non-advertiser bucket the owner can provide and verify his contact information and the more he verifies (along with what we see the server is doing) the better his reputation gets. Remember, spammers' anonymity is their greatest asset.

By the way, I am a part of Reputation Technologies and emailRM.
If you have any more questions I would be more than glad to provide you with as much information as possible.

Thanks,

amujica

RE: Alberto Mujica: Reputation Management for Email | MIT Spam Conference 2007