pesc-org / json-ld-transcript

Creating a machine readable, semantic PESC transcript
https://www.pesc.org/college-transcript.html
1 stars 0 forks source link

Mapping Residency #23

Closed philbarker closed 5 months ago

philbarker commented 8 months ago

I don't see much in CEDS we can use, the closest is residency type which just gives a US-centric view of in/out of state/district for tuition [fees] purposes.

The current model has

(Person) --residency--> (Residency)

(Residency) --pesc:countyCode
                    --pesc:stateProvince
                    --pesc:county
                    --pesc:country
                    --pesc:stateProvinceCode: [defined code]
                    --pesc:countryCode [text]
                    --pesc:residencyStatusCode [text]

Possible solution is to

use ceterms:addressLocality for both pesc:countyCode & pesc:county use ceterms:addressRegion for both pesc:stateProvince & pesc:stateProvinceCode use ceterms:addressCountry for both pesc:country & pesc:countryCode

This merges text and coded values, which may be a problem, if so we could specify that the properties should have structured values that comprise both, e.g.

    "ceterms:addressRegion": {
       "notation": "CA",
       "prefLabel": "California"

Though this kind of breaks CTDL.

BTW ideal solution is to have a URI for the place at, say, "county" level and everything else comes as linked data via that, e.g. "residencyLocation": "https://www.wikidata.org/entity/Q507957", but who has that data to hand.

philbarker commented 7 months ago

Decided on today's call to keep the current properties, maybe merging to use notation and prefLabel as shown above, but with pesc: property not CTDL. Need to write tests.

philbarker commented 7 months ago