---- Ok Hotshot - So how do we adjust the moderation policy?
ronburk - 4:01 pm on May 29, 2010 (gmt 0)
How can we tweak the moderation policy to allow more links - less editing - and more random user satisfaction?
Two-stage moderation. Presence of link moves your post into first stage moderation queue: it's not visible to the masses, but is visible/votable to enough authorized people (probably a larger, less formal group than existing moderators) that someone will likely read it real soon. If just one of those approved people push the NOT OBVIOUS SPAM button, it becomes visible to all, but is actually still removable by vote and has a THIS IS SPAM button visible to a much wider set of mechanically approved (insert tweakable algorithm) "moderators" who really only have the power to push that button. Too many (insert tweakable algorithm) button presses and the thread disappears.
One may also want to prioritize link depth. www.companyname.com is less likely to be turned into a really nasty landing page from hell 2 days after being posted than www.companyname.com/pr/2010/05/27/WebmasterWorld.html.
If done professionally, you won't even realize that a spammer has been in the forums.
Must be assumed to be a human problem, not a technical problem. If it cannot be detected by humans, then we have nothing to talk about. If it can be detected by humans, then an effective response must remove the motivation. E.g., first notify the company that it appears someone is spamming on their behalf. If no help, then publicly post that it appears someone is spamming on their behalf and automatically append a "warning, possible pro spammers work on this product" to any post mentioning the product, etc.
Interestingly, as pro spammers increase in subtlety, they trend towards becoming real content. For example, the "anybody heard of #*$!" is obviously suspect and could be quickly deleted or sequestered to a "product info wanted" forum ghetto. The more subtle "anybody looked at new feature YYY in product #*$!" supplies some actual potential relevant info. If the product is of actual use to WebmasterWorld folks, such spam may result in real followups of useful content. But may also result in "I don't think anybody here uses it, maybe there's something wrong with it." etc. There is a form of the Bonini paradox at work when spammers try to simulate real users more and more accurately. Eventually, they have the actual benefits of real users.
In any case, there's sort of a decision point of whether or not you wanted to move into mass moderation or not. Is the the problem worth the coding/maintenance effort that would cost?