Now that we have data indexed, we need to build a search interface. My original model had been to use the mechanism shown in the ChatGPT for Genetics approach, but I am not wedded to this idea and I've since seen suggestions of other approaches. Using knowledge graphs with LLMs is a hot area right now, and there seem to be quite a few ways of approaching it.
To use the ChatGPT+GraphQL approach we'd want to:
(1) build a GraphQL request router FoxGraphQL Integration
(2) build a front end interface to interact with that request router (which is what ChatGPT for Genetics does.)
(3) Start testing queries and then building additional mechanisms to support additional queries.
I'm not excluding other search interface possibilities here either. Some queries could likely be done using ArangoDB tools. For example, they have a search interface, see ArangoDB Search
Now that we have data indexed, we need to build a search interface. My original model had been to use the mechanism shown in the ChatGPT for Genetics approach, but I am not wedded to this idea and I've since seen suggestions of other approaches. Using knowledge graphs with LLMs is a hot area right now, and there seem to be quite a few ways of approaching it.
To use the ChatGPT+GraphQL approach we'd want to:
(1) build a GraphQL request router FoxGraphQL Integration (2) build a front end interface to interact with that request router (which is what ChatGPT for Genetics does.) (3) Start testing queries and then building additional mechanisms to support additional queries.
This is an important but challenging bit of work.