Chapter 13. Deep Learning for Text

Until now, we have explored the use of machine learning for text in a variety of contexts – topic modeling, clustering, classification, text summarization, and even our POS-taggers and NER-taggers were trained using machine learning. In this chapter, we will begin to explore one of the most cutting-edge forms of machine learning –Deep Learning. Deep Learning is a form of ML where we use biologically inspired structures to generate algorithms and architectures to perform various tasks on the text. Some of these tasks are text generation, classification, and word embeddings. In this chapter, we will discuss some of the underpinnings of deep learning as well as how to implement our own deep learning models for text. Following are the topics we will cover in this chapter:

• Deep learning
• Deep learning for text
• Text generation