In the previous chapter, you saw NLP applied to the construction of knowledge graphs to support semantic search over collections of things (articles, products, documents, and so forth).
This relied on an NLP task called entity extraction, or NER.
But NER is only one of the ways in which NLP can interact with knowledge graphs, out of three broad categories:
Where knowledge graphs are populated with entities, facts and knowledge come from applying NLP techniques to natural language text.
This includes what you learned in Chapter 12 as well as the cases where NLP is used for fact extraction to build question-answering knowledge graphs.
This category is natural language as input to a knowledge graph.
You will see in this chapter how natural language can also be generated from knowledge graphs.
This can be used to produce a conversational answer to a query or for automated report generation, as some examples.
This category is natural language as output.
Somewhere in between, knowledge graphs can be a tool providing structured context for NLP tasks, either at the lexical or at the conceptual level (or both).
Figure 13-1. Types of interaction between natural language and knowledge graphs
This chapter is split into three parts. In the first part, you will learn how NER can be complemented with ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month, and much more.
