In this paper, http://www.cs.cmu.edu/~leili/pubs/lu-baylearn2015-twinet.pdf, the authors look at the problem of recognizing entities in natural language. This is often the first step in question answering, conversations, and a host of other NLP use cases. For a sequence of text tokens, a named entity recognizer identifies chunks of tokens that belong to a predefined category of persons and organizations.
Named entity recognition using a twisted neural network
Get Neural Network Programming with TensorFlow now with O’Reilly online learning.
O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.