Support entity-entity relationships

Now that we have a base interface for editing entities, we can more easily curate their basic data and potential relationships between them. Following up @pdcp1's mock here, we can now start to implement basic relationships, starting with a catch-all related relationship while we figure out a more durable system to commit to editing.

The primary function this would serve is to relate our soup of alleged developers, deployers, and victims implicated in AI incident reports.

Additionally, an entity relationship system can allow us to record and name AI systems and models and relate them to incidents, developers, and other AIID data.

Per https://github.com/responsible-ai-collaborative/aiid/issues/2536#issuecomment-1960443680:

An initial mockup for the Entity edit page is in our Figma workspace https://www.figma.com/file/KI28jWrOO3soKp9dTg7b0d/Entities-workflow?type=design&node-id=0%3A1&mode=design&t=a1M9nWWeHzoisF2d-1

We should take this as a kick-off design. I added some fields that I consider appropriate but we can discuss which ones are more important to implement in an initial phase. Feel free to edit it. I'm open to any suggestions and discussions.

Option 1

Simply extend the entity object with an array of related entity IDs.

{
  "entity_id": "amazon-warehouse-workers",
  "name": "Amazon warehouse workers"
  "related": [
      "amazon", // entity_id
       "warehouse-workers"
  ]
}

Pros:

Simpler at first. Defers having to define further what "related" means.
Finding related entities does not require a secondary lookup in linked data.
Easy to migrate from later on.

Cons:

related is a symmetric relationship; we would want both entities to have the property. Thus, we would have to synchronously update two entities that are currently stored separately in the entities collection – and ideally validate this continuously.
Further validation: entities and their IDs in relationships must exist and be well-formed. We currently lack validation over entity data.
Flexibility and hard-coding. It's relatively easy to expand/migrate from this implementation later on, but we are hard-coding the field "related" onto an entity. When we wish to include another entity-entity relationship (e.g. developed_by), will we just add another field to the mongo document? Or begin the migration to....

Option 2

Create a collection specifically for entity-entity relationships, i.e. an edge table. We could store these as basic semantic triples (subject, object, predicate), with additional metadata about whether or not the property/relationship is symmetric.

We would be entering an early implementation of semantic triples: https://en.wikipedia.org/wiki/Semantic_triple

{
  "pred": "related",
  "sub": "amazon-warehouse-workers",
  "obj": "amazon",
  "is_symmetric": true
}

Pros

Start advancing towards semantic and linked data. :-)
More flexible if we wish to start adding other property relationships already in our data, e.g. developed_by.
Easier to start adding other relationships (e.g. developed_by, which would be complemented with a develops relationship).
Easier to start adding metadata on the relationship (e.g. status, confidence, symmetry, provenance).

Cons

Opening up the linked data storm!
Lookups become more complicated, both in static and dynamic contexts. This would have implications for the GraphQL API we are developing.
Validation: As with above, the ideal implementation has a way to validate that all edges/relationships are valid, i.e. connect two well-formed and existing entities with valid IDs.

Other options

MongoDB purports to have useful graph-like traversal qualities that we could explore using directly: https://www.mongodb.com/resources/basics/databases/mongodb-graph-database

Nota Bene

We are casually entering further into the territory of linked/semantic data. Many implementations and standards for such data already exist, e.g. RDFa, JSON-LD.

We should consider how we can make existing standards of linked data work for us in the future. Incident data and reports of AI harms are inherently graph-like in the real world as well as on AIID.

responsible-ai-collaborative / aiid