Summary
In this chapter, we addressed the process of preparing data and discussed pipelines. We illustrated several techniques for extracting text from HTML, Word, and PDF documents. We also saw how Apache Tika can be used easily with any kind of document for extraction. We showed that a pipeline is nothing more than a sequence of tasks integrated to solve some problem. We can insert and remove various elements of the pipeline as needed. The Stanford pipeline architecture was discussed in detail. We examined the various annotators that can be used. The details of this pipeline were explored, along with how it can be used with multiple processors. In next chapter, Chapter 12, Creating a Chatbot we will work on creating a simple chat bot to demonstrate use of NLP we have seen so far.