Ask HN: Rate this idea. [Interactive Job Interview Practice.]
Here's the concept: machine learning and natural language processing deployed for the purpose of simulating and grading job interviews. Later developments could be speech processing (sentiment and tone) and, much, much later, visual processing for the assessment of social skill.
This would serve the same purpose as mock interviews, but it would give people the opportunity to try different approaches and see how they compare. It would also allow people to calibrate the type of interview (how technical, how aggressive, how informal) according to what they expect when they face the real thing.
Automated essay grading is already semi-solved, if not solved (by "solved", I mean in agreement with human graders as often as 2 humans agree). Bayesian techniques could be used to assess answers for positive and negative tone, just as they're used in spam detection. So I feel like the pieces are there. It may be tough to grade answers to "What is your biggest weakness?" in a way that corresponds to human reaction, but I think it's a solvable problem.
Getting data would be the hardest part of the problem, in my opinion, since this is the sort of thing that people care to keep confidential. That, I think, is the biggest weakness of this idea. Data could be "crowd-sourced", but something tells me that the people who actually make hiring decisions wouldn't be in that "crowd".
I think, and I could be wrong, that such a product could be immensely useful. It'd probably be a long time before the AI is as effective as a human career counselor, and first iterations (which would be text-based, without voice-processing) would ignore social skills, the most important component of a job interview. Still, it would be a start.
Is anyone else interested in this? I see a lot of potential here, or it could just be an impractical pipe dream. Any thoughts? Has something like this been tried before?