This and TipJoy have inspired me. What about making a social news site, a Wikipedia-like site, or a classified ads site where you can read and search all you want for free, but you have to make a micropayment to post? Micropayments would be in "credits" that cost 4.5 cents each. $5 gets you 100 credits. (There's a 50 cent processing fee.) You could also write the Terms of Service such that spamming forfeits your $5.
You could even combine this with a TipJoy-like idea and have commenters tipped for every up-vote. (Down-votes would subtract from a comment score, but not affect the tip.)
This would kill off the spammers. Paying $5 per spam post would ruin their business model. I suspect that it would also increase the quality of the posts if posters could earn money from their efforts.
We really need a micropayments protocol. I've been working on one intermittently for about 3 years. Please let me know if you'd like to talk more about it.
I think it will take a simple protocol, like SMTP, using one-time passwords to authorize transfers. Your software will use transfer services to exchange money at different banks, streaming the money to minimize theft.
why would someone pay you money to submit information to your wikipedia? I think Wikipedia got that portion right, users can submit all they want, but you have a bunch of thankless moderators filtering your content for you
Maybe Wikipedia wouldn't work unless you made the whole service "for-pay" in some way. The nice aspect of requiring users to pony up some amount, which then becomes a pool that you make micro-payments from. So you get the gating aspect of a for-pay service combined with the market forces resulting from the micropayments. And yet, the users are really not paying the $5 up front.
Wikipedia might well not work, simply because the free one works so well.
I disagree. Spam has a low success rate, so it's only extremely profitable when the cost per advertisement is zero. If you have a service where it costs _any_ money to post then spammers will go find somewhere else.
I think it could work well for classified ads since that's something people are willing to pay for. Not so well for wikipedia or social news, because who cares enough to pay money for those?
I'd price classified ads at one dollar each -- small enough that it's practically free, but large enough that you could easily make a lot of money out of it if people ever decide to abandon craigslist.
Would they be willing to pay 50 cents per post? Would they be willing to pay $5 per post? You could write the TOS for such a site that spamming forfeits your $5.
I think sites that run some kind of an automatic ranking or otherwise a user-driven voting system are better protected from spam. To make it even better there should the second level with weighted users (user ranking in other words) to compensate massive automatic registrations of fake users.
I'm sure it is much easier for Flickr, for instance, to fight spam than for Cragislist, because Flickr has an excellent automatic ranking system of photos based on users' (implicit) votes plus users' weights. This at least prevents spam from slipping into the daily top-500, and the rest is not that important. They host all kinds of garbage anyway and that doesn't bother the community.
This is what makes spammers turn away from Flickr, I think. I have never seen any spam on Flickr's top-500 page (well, except those crappy images with anything-Apple insanely upvoted by a bunch of Apple fans).
The problems I see with a Bayesian classifier in this application are:
1. Spammers get instant feedback on whether their post was flagged. When they send me email spam, for example, they have no idea how badly they're doing.
2. Spammers have access to a large ham database to draw from -- all the posts that were not filtered. With email, they never see any user's highly individualized ham database.
3. Bayesian classifiers require a significant amount of manual intervention, especially given 1 and 2. I'm happy to correct classification errors in my email when they happen at a rate of roughly 1/200,000, but if I had to stay on top of dozens of false negatives and false positives every day, I just wouldn't. Very quickly the database would become too noisy to be of much use.
You could even combine this with a TipJoy-like idea and have commenters tipped for every up-vote. (Down-votes would subtract from a comment score, but not affect the tip.)
This would kill off the spammers. Paying $5 per spam post would ruin their business model. I suspect that it would also increase the quality of the posts if posters could earn money from their efforts.