search

designforcontext / aac_review_tool

MIT License
5 stars 7 forks source link

AAT Description (activity) #57

Open VladimirAlexiev opened 7 years ago

VladimirAlexiev commented 7 years ago 
<http://data.thewalters.org/object/155/description>
  crm:P2_has_type <http://vocab.getty.edu/aat/300404670>

aat:300404670 is Description, but the activity of describing something: analytical functions, <functions by general context>, functions (activities), Functions (hierarchy name), Activities Facet.

Need to pick something better

workergnome commented 7 years ago

I think you're referring to aat:300080091. With the fuzziness of AAT + CRM:P2, I don't know that we can be militant about this, but if you have a better AAT suggestion, I'd add it to a future version

VladimirAlexiev commented 7 years ago

Asked Patricia to make one: https://jira.getty.edu/browse/ITSLOD-508. (I'm sure she'd take exception to your calling AAT "fuzzy")

azaroth42 commented 7 years ago

This is the challenge of any vocabulary without an ontology, and then trying to align a very opinionated ontology with it.

If we were to cast AAT into CRM, aat:300080091 would probably be a subclass of Production (and perhaps actually called 'describing'), and it would produce a Description, subclass of Linguistic Object (or similar). AAT isn't fuzzy on its own as it is a hierarchy of strings, but it is fuzzy when used in conjunction with ontologies.

VladimirAlexiev commented 7 years ago

You need P2 for a conceptual object (Linguistic Object). So you need an AAT concept representing object, not activity. Currently there isn't one but Patricia will add one. Nobody's trying to align 45k AAT concepts to 90 CRM classes.

VladimirAlexiev commented 7 years ago

Patricia made aat:300411780 descriptions (documents)

workergnome commented 7 years ago

Great. I'll add that in the next version

VladimirAlexiev commented 7 years ago

example for test: http://data.americanartcollaborative.org/page/wam/object/33060/description

workergnome commented 7 years ago

I've got a note to see if I can convince ISI to update all their mappings, but it seems unlikely that they're going to do so. However, I've also flagged it in https://github.com/linked-art/linked.art/issues/91, to update for the future.