@stuartyeates @egonw @wdscholia,
-
Kingsley Uyi Idehenwrote last edited by [email protected]
@stuartyeates @egonw @wdscholia,
Given the example at: https://github.com/WDscholia/scholia/blob/51a2d05f9f857162f3f75780a68ddfe8769c8fb0/scholia/app/templates/work_data.sparql#L5
It would be helpful if you include comments and sample values at the start of the query code. You’ve already done a significant portion of the work for integration into an LLM using tools like #OPAL (our knowledge graph bridge to #GPT language models).
-
Egon Willighagenreplied to Kingsley Uyi Idehen last edited by
@kidehen okay, what I hear is that it is something that we already knew: it takes work. you do not have to explain this to me
-
Kingsley Uyi Idehenreplied to Egon Willighagen last edited by
My point is that adding a comment with sample placeholder variables will streamline collaboration.
For instance, I can quickly create a predefined prompt:query template for LLM processing that’s easy to demonstrate. This would show others how they can use Scholia and contribute to its development.
You can start with the example I mentioned earlier.
-
@egonw @kidehen yes also see https://github.com/sib-swiss/sparql-examples/ @Andrawaag was working on setting up the same idea for wikipathways. A repository of query examples that can help any intelligence get more value out of open data.