Deep Learning Illustrated: A Visual, Interactive Guide to Artificial Intelligence

11. Natural Language Processing

In Chapter 2, we introduced computational representations of language, particularly highlighting word vectors as a potent approach for quantitatively capturing word meaning. In the present chapter, we cover code that will enable you to create your own word vectors as well as to provide them as an input into a deep learning model.

The natural language processing models you build in this chapter will incorporate neural network layers we’ve applied already: dense layers from Chapters 5 through 9, and convolutional layers from Chapter 10. Our NLP models will also incorporate new layer types—ones from the family of recurrent neural networks. RNNs natively handle information that occurs in sequences such as natural ...