Ingesting data into a knowledge graph

There is a lot to consider before jumping straight into creating a knowledge graph from our cleaned abstract data. As with previous chapters, we must consider the structure of the graph we are aiming to produce first. We will then need to process our abstracts to extract terms of interest. Then, once we have terms, we can create a list of edges to import into igraph.

Getting the ingestion right into the knowledge graph is crucial and this all stems from how you conceptually and practically design your graph schema. The following section shows how to design your schema to make sure your knowledge graph works the way you expect it to.

Designing a knowledge graph schema

Before jumping straight into data ingestion, we must consider the structure of our knowledge graph. For our use case, we’re interested in connecting related documents and concepts.

In terms of nodes, we have both abstracts and terms. Our abstracts have only an ID...