W3C

– DRAFT –
FHIR RDF

16 September 2021

Attendees

Present
David Booth, Emily Pfaff, Gaurav, Gopi, Rob Hausam
Regrets
-
Chair
David Booth
Scribe
dbooth

Meeting minutes

Issue 77

https://github.com/w3c/hcls-fhir-rdf/issues/77#

Discussing 4a vs 4b

Gopi: Suggestion changing birthDateObject to birthDateExtension.

david: good idea

AGREED: Prefer 4b over 4a

David: 3b vs 4b?

gaurav: prefer 3b

emily: agree

gopi: agree

AGREED: Prefer 3b over 4b

david: Compare 5a vs 3U

emily: Don't like the extra ont relationships.

gopi: still prefer 3b over both of these

gaurav: calling them extensions might be simpler.

gaurav: prefer not to create all these new properties

gaurav: The fhir:Extension in 3bU seems like it could be fhir:Boolean

david: Or it could be both a fhir:Extension and a fhir:Boolean.

David: 5a, 3U or 3bU vs 3b?

emily: Prefer 3b over all of these.

gopi: 3bU makes sense if you are adding a certainty.

david: We should probably get eric's input before eliminating these, to hear his perspective.

gaurav: Slightly prefering 3bU over 3b.

gopi: One advantage of 3bU is when you're entering data you can just use that new predicate directly.

david: Not sure of the impact of 3gU (and other RDF-style options) on conversion from RDF back to JSON. Assuming that the standard machinery would not know about the extension. Seems it would at least require converting the extension URL from an RDF node to a string.

rob: might be able to pull in a package to tell the machinery about the extension.

Action: Gaurav to spin up on JSON-LD framing

david: more ideas to explore for these options?

gopi: What about data validation?
… Need to validate cases, using SPARQL queries.

david: Yes, important to consider impact on validation.

gopi: Been looking at RDF reification. Extensions are similar in my mind, for the Certainty example.

david: For the fhir:active example, RDF reification would be applicable, because "Certainty" is a statement about the boolean statement. But I don't think reification would be applicable in other extension examples, such as the birthDate example, which is just a matter of adding more information (time of birth).

ADJOURNED

Summary of action items

  1. Gaurav to spin up on JSON-LD framing
Minutes manually created (not a transcript), formatted by scribe.perl version 136 (Thu May 27 13:50:24 2021 UTC).

Diagnostics

Succeeded: i/Suggestion changing/https://github.com/w3c/hcls-fhir-rdf/issues/77#

Succeeded: i/Suggestion changing/Discussing 4a vs 4b

Succeeded: i/avid: 5a, 3U/david: Or it could be both a fhir:Extension and a fhir:Boolean.

Succeeded: i/ADJOURNED/david: For the fhir:active example, RDF reification would be applicable, because "Certainty" is a statement about the boolean statement. But I don't think reification would be applicable in other extension examples, such as the birthDate example, which is just a matter of adding more information (time of birth).

Succeeded: i/gopi: Been looking/david: Yes, important to consider impact on validation.

Succeeded: i/rob: might be able/david: Not sure of the impact of 3gU (and other RDF-style options) on conversion from RDF back to JSON. Assuming that the standard machinery would not know about the extension. Seems it would at least require converting the extension URL from an RDF node to a string.

No scribenick or scribe found. Guessed: dbooth

Maybe present: AGREED, david, emily, rob