search

vivo-community / vivo-graphql

graphql endpoint for the vivo scholar project
BSD 3-Clause "New" or "Revised" License
0 stars 1 forks source link

General overall data model discussion #12

Open outten45 opened 5 years ago

outten45 commented 5 years ago

Set up a data model discussion with TAMU. Data model comparison and defined outside of code (for example, in JSON format).

Initial Questions for TAMU

chemGoggles commented 5 years ago

Hi @delsborg @roflinn, I got a chance to review the API documentation today. My first impressions are that overall it looks similar. That said...

  1. The Product Evolution data spec has a lot more going on, particularly when it comes to date resolution, roles, and dates on roles.

  2. People tend to be referenced in the data model with their organization ID and organization Label. For example, you can traverse from Documents (publications)-->authors-->organizations. Same thing for keywords. Not sure how useful that is since a person's publication or keywords doesn't necessarily relate to every department with which they have an affiliation.

  3. Still thinking about sorting options and the possibility to tag information as "featured"

It seems like with the flexibility of GraphQL, there's no downside to putting a ton of data into the data model...maybe I'm wrong on that but seems like the safest option so that we don't run into limitations down the road.

Damaris

chemGoggles commented 5 years ago

@roflinn will look into setting up a preliminary data discussion with TAMU. Should we document our questions for TAMU in the description section of this issue?

delsborg commented 5 years ago

A couple of pre-req tasks:

  1. Create a branches of data model for TAMU, PE Spec, and Original Duke GO model. TAMU models now at: https://github.com/TAMULib/Scholars/tree/sprint3-experiment/middleware/schema/base
  2. Reconcile the models