I am a software developer working at INRAE (French national institute for agroecologics). I work since 4 years about subjects like ontology (Knowledge graph) and information system. I use the community edition since almost this time for various purpose, and i am here because i have questions about the graphql library ! (which i would love to use as an automated generated graphql CRUD based on owl/rdf ontology)
Me and my friend are doing something similar. Please tell me more.
I don't really know on what to elaborate, but I will give more context ^^.
As I said, I work in a team (french public research) working on semantic web and ontologies. The labs in question devloped a software called Opensilex, which is an information system for scientific data and object, based on ontologies. It uses a triplestore to store data accordingly to user defined ontologies (whereas those ontology need to use/extend an opensilex ontology used to described data, tracability, follow FAIR spec, etc...). There is multiple (between 7-10) instances of this software for different data and research labs, a lot involving vine and crops, serving the data of phenotyping platform for example.
But while I work here (or closely to them) for a while now, a big source of slowdown/bug is the fact that the software use a REST api... But all the services are written by human hands, and as it is a french research labs, mostly by beginners which start their professional life there before living, as the public research is not very attractive in term of salary and situation (not a lot of permanent contract), especially for software ingineers.
I had the opportunities to work on a project with agronoms where I was alone making the software and modeling decision, and I then used neo4j to store knowlegde graph. Then, I came back on the opensilex dev team, and, while it is extremely relevant to work with ontologies in OWL/RDF format, in does not appear to me that storing big data in a triplestore is wise, especially for CRUD features, and morever now that librarie like graphql neo4j exists now.
I am currently developing on my spare time a prototype of an autogenerated CRUD from owl => gql definition => neo4j-raphql + appollo server and... It's working for the core fonctionalities.
Anyway, I talked a lot here, but what you guys do ? Maybe we set up a more formal exchange ?