Wonderful work! I've been very curious to see a neural network approach to this problem. You mention the CRF work by McCallum. How does your method compare to that, which seemed like the previous state of the art to me?
Wonderful work! I've been very curious to see a neural network approach to this problem.
You mention the CRF work by McCallum. How does your method compare to that, which seemed like the previous state of the art to me?