Introducing knowledge graphs
In complex fields, such as science and medicine, the sheer amount of data and literature available on specific topics is hard to overstate. The same goes for knowledge management in established companies and industries where, over time, institutional knowledge in the form of textual information builds up, becoming too large to sensibly disseminate. In both of these cases, a knowledge graph may help to alleviate issues associated with too much disparate information.
The aim of a knowledge graph is to link together related information, text, and documents in a sensible and searchable way.
In the case of knowledge graphs using text, links in a graph often represent related documents or articles. Text processing and NLP are huge fields in themselves, so for the purposes of this chapter, we will be keeping methods for working with text simple. Of course, the quality of text data has a large impact on the preparation of data for knowledge graph ingestion...
