Kevin here. No, it's not perfect and never will be. I definitely use a bayesian classifier in our process among other things. That said, if you see mistakes, p

Kevin here. No, it's not perfect and never will be. I definitely use a bayesian classifier in our process among other things. That said, if you see mistakes, please flag them! My system does get smarter with more data, so when people flag mistakes for us to correct, it makes a difference both immediately and in the long run. If you have any other thoughts/critiques/ideas please feel free to let me know :)