There are many natural language applications that require machine learning and deep learning techniques. We’ll discuss some of the following in our machine learning and deep learning chapters:

• Answering natural language questions—For example, our publisher Pearson Education, has a partnership with IBM Watson that uses Watson as a virtual tutor. Students ask Watson natural language questions and get answers.

• Summarizing documents—analyzing documents and producing short summaries (also called abstracts) that can, for example, be included with search results and can help you decide what to read.

• Speech synthesis (speech-to-text) and speech recognition (text-to-speech)—We use these in our “IBM Watson” chapter, along with inter-language text-to-text translation, to develop a near real-time inter-language voice-to-voice translator.

• Collaborative filtering—used to implement recommender systems (“if you liked this movie, you might also like…”).

• Text classification—for example, classifying news articles by categories, such as world news, national news, local news, sports, business, entertainment, etc.

• Topic modeling—finding the topics discussed in documents.

• Sarcasm detection—often used with sentiment analysis.

• Text simplification—making text more concise and easier to read.

• Speech to sign language and vice versa—to enable a conversation with a hearing-impaired person.

• Lip reader technology—for people who can’t speak, convert lip movement to text or speech to enable conversation.

• Closed captioning—adding text captions to video.