The difference between JSON-LD documents and Hypergraph objects

I’m building a Hypergraph platform with JSON-LD interfaces and bring it to the world. The platform helps technical professionals build knowledge graphs that work similarly to object databases, but with a knowledge graph representation, and where each commit transaction ensures that the graph is logically consistent.

While building the platform, I had to think a lot about the nature of things, and finding the right wording for what is processed in the platform. What is interesting is that the correct wording hides in plain sight, but needs a deep understanding to get right.

Generated by the author with Dalle-3

Two key words kept me engaged for a long time, what to call the instances of types in the platform, and the representations of them outside of it. Should they be:

• records, like in a database,
• objects, like in Javascript and object-oriented languages like Java,
• instances of types, like in information modelling, or
• documents, as they are seen by the developer and API users?

I have had other words come up too, but these were the main ones. This question hinged on getting a deep understanding of how the platform works. Additionally, in a database such as TerminusDB, there is a notion of subdocuments, that are essentially…