Natural language generation (NLG), which is a sub-field of artificial intelligence, is a natural language processing task of generating human-readable text from various data inputs. It is an active area of research that has achieved great popularity in recent times.
The ability to generate natural language through machines can have wide variety of applications, including text autocomplete feature in phones, generating the summary of a document, and even generating new scripts for comedies. Google's Smart Reply also uses a technology that runs on similar lines to give reply suggestions when you're writing an email.
In this chapter, we will look at an NLG task of generating a book script from another Packt book that goes by the name of Mastering PostgreSQL 10. We took almost 100 pages of this book and removed any figures, tables, and SQL code. The data is fairly large and has enough words for a neural network to learn the nuances of the dataset...