Re:Quora pays thousands for posting questions
by fourthark on Sunday December 23, 2018 @02:04AM
(#18744861)
I don't know about the vetting process, but the ranking is by popularity not accuracy.
So, I like this theory but I doubt the quality of what your algorithms would learn this way.
Re:Quora pays thousands for posting questions
by ChuckMcM on Sunday December 23, 2018 @02:38PM
(#18747883)
If you look at it from a dialoging system point of view, it still works. What would give someone the impression that a dialog system was "good" ? Answer, they like the answers it gives. So knowing "popular" answers to questions can give you two vectors, both sentence construction (which is to say how the answer should be presented) and accuracy score (which is how much does popularity depend on accuracy).
The canonical example is "Siri, where can I hide a dead body?" This is a question where being accurate will leave the consumer less satisfied than with giving a "popular" answer.
I didn't appreciate how hard it was to come up with good QA pairs for a topic until IBM Watson acquired the company I was working for (Blekko) and got to see how things got made. I isn't something you can farm out to Mechanical Turk, and without AGI it isn't something that is easily auto generated from a knowledge graph. Ten thousand questions about the insurance business and vetted and ranked answers? Solid gold.