the-fair-cookbook
the-fair-cookbook copied to clipboard
FAIRplus
Issue on page /content/recipes/interoperability/creating-data-dictionary.html
A data dictionary can be very useful. The columns suggest are reasonable. But would it be better to use .obo format? The information that has to be captured fits pretty well in .obo, in my experience.
Dear @chrishmorris , sorry for taking so long to come back to you.
Thanks, first of all, for your suggestion! You are certainly right that obo might be used as a conceptual frame for data dictionaries!
We (i.e. FAIRplus), but better: The authors developed this specific recipe with hands on a specific IMI project, in this case ABIRISK. In ABIRISK, the data dictionary was provided in the described form -- and, speaking for myself, I am actually already very happy that there was a data dictionary at all!
I nevertheless agree that it would be a good addition to describe how to implement a data dictionary with obo. We (FAIRplus) are constrained with resources (like everyone... 😉 ), so I cannot promise any action from our side.
Two proposals:
- I bring it up in our "Squad" =team discussion to see if someone of us (FAIRplus) has the capacity to dive into obo and write down a recipe.
- Maybe you have time yourself to describe how an obo data dictionary would work? If so, I would be happy to assist in getting your thoughts into the cookbook!
What do you think?
(pinging @daniwelter @weiguUL @proccaserra as authors of the recipe, just for your reference, no action needed)
discussed during bookdash #10 more discussions with the authors: positive points: use of a defined file format for structuring information negative points: less straightforward to use, specific tooling for editing
Data modelling would be a nice follow-up discussion to have in keeping with building a data dictionary. The question to @chrishmorris is whether the suggestion is to use obo format to record the variables declared in the data dictionary or to build a data model relating the variables defined in the data dictionary
AI: discuss possibility of recipe on building data models
A data dictionary is a list of column names / variables, preferable with definitions, units, and constraints. It seems to me that OBO can help to communicate this information.