TypeKey and the Trust Web
After browsing Movable Type, I read that they have created TypeKey. TypeKey is, from the looks of it, a single sign on for blogs.
TypeKey primarily has been created to fight comment spam appearing in blogs.
I ponder how the semweb could help with this problem. There has been an effort to generate a trust web using semantic web technologies. Can this web be used to help fight comment spam?
My guess, though untested, would be yes. If blog owners require sign in before posting a comment, then an identity is collected. My blog system can then query the semantic web and get a trust metric for the ID.
TypeKey only provides authentication services. It leaves authorization up to the blog backend. This is a wise choice, and it leaves open an implementation for authorization based on the semweb's trust web.
I can see a authorization system based on the trust web, where users have certain trust metrics for posting to blogs. Note that a blog posting trust metric would be different than a "take care of my grandma" trust metric, so there must be a distinction there.
As users, once authenticated, post to blogs without generating (subjectively measured) negative feedback, their intrinsic trust metric increases. For instance, if a user that has posted 10000 posts across the web with no spam, then that user can relatively be trusted to post to your blog without harm.
How would an authorization agent know the blog posting trust metric of an authenticated user? Possibly by querying a trustkey authorization service. It could assign a blog posting trust metric and send it back to the blog backend.
It would be interesting to investigate if the information of "User X has posted a non-harmful blog comment" from the web. RSS feeds, to my knowledge, do not include the comments (though nothing technical would stop this from happening). So getting this information as triples would be more difficult than merely harvesting statements from the semweb.
It sounds good, and an interesting application of the trust web. Can this be built? More importantly, would people use it?
TypeKey primarily has been created to fight comment spam appearing in blogs.
I ponder how the semweb could help with this problem. There has been an effort to generate a trust web using semantic web technologies. Can this web be used to help fight comment spam?
My guess, though untested, would be yes. If blog owners require sign in before posting a comment, then an identity is collected. My blog system can then query the semantic web and get a trust metric for the ID.
TypeKey only provides authentication services. It leaves authorization up to the blog backend. This is a wise choice, and it leaves open an implementation for authorization based on the semweb's trust web.
I can see a authorization system based on the trust web, where users have certain trust metrics for posting to blogs. Note that a blog posting trust metric would be different than a "take care of my grandma" trust metric, so there must be a distinction there.
As users, once authenticated, post to blogs without generating (subjectively measured) negative feedback, their intrinsic trust metric increases. For instance, if a user that has posted 10000 posts across the web with no spam, then that user can relatively be trusted to post to your blog without harm.
How would an authorization agent know the blog posting trust metric of an authenticated user? Possibly by querying a trustkey authorization service. It could assign a blog posting trust metric and send it back to the blog backend.
It would be interesting to investigate if the information of "User X has posted a non-harmful blog comment" from the web. RSS feeds, to my knowledge, do not include the comments (though nothing technical would stop this from happening). So getting this information as triples would be more difficult than merely harvesting statements from the semweb.
It sounds good, and an interesting application of the trust web. Can this be built? More importantly, would people use it?