Fact check and tag at the same time, as a contract business?
12/13/09 10:05 Filed in: indexing
Paul
Spinrad has an interesting idea for a business
here:
Fact-Checkers and Certified Public Logicians
It's fantastic that so much written knowledge is becoming generally accessible and cross-linked these days, but this is just an intermediate stage-- a universal library on the way to becoming a universal brain. The missing piece is encoding the underlying meaning of the stored text, the deep-structure logic behind it. It's one of the oldest challenges in Computer Science, and there has been lots of progress and companies dedicated to doing this. Powerset, for example, has software that has parsed and can answer questions from all of Wikipedia.
The thing is, you really still need a person to get it most reliably right, because people understand the way the world works. Luckily, we already have people whose job is very close to doing this already-- they're called fact-checkers or researchers, and they work for every reputable publication.....
I have wondered for years, as magazines, newspapers, and other news organizations have been hemorrhaging money and employees, why someone hasn't gone into the contract fact-checking business. Like, it could be an extension of Snopes.com. There's a huge redundancy in every publication having their own research desks, so they could lay off all of their fact-checkers and then outsource the job to the new, independent company that the best of them then all go to work for. Meanwhile, the company could also be hired by anyone else. Then, when the public sees the "Fact-Checked by MiniTrue (SM)" seal on someone's independent blog, they know the information there has the same credibility as the big boys.
Now, what if these fact-checkers didn't just vet and correct the text? While they dig into the logic and accuracy of everything, as usual, they could also use some simple application to diagram the sentences and disambiguate the semantics into a machine-friendly representation. Just a little extra clicking, and they could bind all the pronouns to their antecedents, and select from a dropdown box to specify whether an instance of the string "Prince" refers to the musician Prince or to Erik Prince-- the president of XE, the company formerly known as Blackwater-- within an article that for whatever reason mentions both of them.
Then you would really have something. The text wouldn't just be fact-checked; its underlying meaning could be added into a shared pool of human knowledge, chained through, verified or denied, and used in other ways by any technology that may now exist or may exist in the future.
Fact-Checkers and Certified Public Logicians
It's fantastic that so much written knowledge is becoming generally accessible and cross-linked these days, but this is just an intermediate stage-- a universal library on the way to becoming a universal brain. The missing piece is encoding the underlying meaning of the stored text, the deep-structure logic behind it. It's one of the oldest challenges in Computer Science, and there has been lots of progress and companies dedicated to doing this. Powerset, for example, has software that has parsed and can answer questions from all of Wikipedia.
The thing is, you really still need a person to get it most reliably right, because people understand the way the world works. Luckily, we already have people whose job is very close to doing this already-- they're called fact-checkers or researchers, and they work for every reputable publication.....
I have wondered for years, as magazines, newspapers, and other news organizations have been hemorrhaging money and employees, why someone hasn't gone into the contract fact-checking business. Like, it could be an extension of Snopes.com. There's a huge redundancy in every publication having their own research desks, so they could lay off all of their fact-checkers and then outsource the job to the new, independent company that the best of them then all go to work for. Meanwhile, the company could also be hired by anyone else. Then, when the public sees the "Fact-Checked by MiniTrue (SM)" seal on someone's independent blog, they know the information there has the same credibility as the big boys.
Now, what if these fact-checkers didn't just vet and correct the text? While they dig into the logic and accuracy of everything, as usual, they could also use some simple application to diagram the sentences and disambiguate the semantics into a machine-friendly representation. Just a little extra clicking, and they could bind all the pronouns to their antecedents, and select from a dropdown box to specify whether an instance of the string "Prince" refers to the musician Prince or to Erik Prince-- the president of XE, the company formerly known as Blackwater-- within an article that for whatever reason mentions both of them.
Then you would really have something. The text wouldn't just be fact-checked; its underlying meaning could be added into a shared pool of human knowledge, chained through, verified or denied, and used in other ways by any technology that may now exist or may exist in the future.