buda-base / owl-schema

BDRC Ontology Schema
11 stars 2 forks source link

alternative way of recording role #92

Open eroux opened 5 years ago

eroux commented 5 years ago

Currently we have only one pattern possible to record a role that's associated with a person: :personEventRole. But in some cases (like the 4 first Dalai Lama or the 2 first Karmapa), the roles were assigned to people in a retroactive way and so there never was an event when they assumed office. Yet the information that they are designated as having this role seems worth recording.

We could also record the number they were assigned in a semantic way. Currently this is recorded only in the OfficeTitle names (ex: tA la'i bla ma 13 for P197).

eroux commented 5 years ago

I understand what's going on a bit better now. There is a missing information in the roles though, that should be added in the ontology (and in the migration) one way or the other. The information is that a Role is or is not a tulku line. The current website currently handles it with the conditional:

where contains(data($i/r:description),'skye brgyud')

in xq:lineage-offices() of modules/gwt/web-content-generator-persons-tulkus.xq. I think this information should be recorded more semantically. There could be, for instance a subclass of :Role that would be :TulkuLine, something like that?

eroux commented 5 years ago

Now, a problem with Tulku lines is that the first of the line (like P80) has a strange event:

<p:event circa="?" type="assumeOffice">
    <p:office pid="R8LS12876">tA la'i bla ma'i sprul sku'i skye brgyud/ </p:office>
</p:event>

which doesn't make much sense... so maybe :TulkuLine could have an associated property :tulkuLineFirst. I think it would make more sense... wdyt?

xristy commented 5 years ago

We currently have :PersonAssumesOffice and :PersonLeavesOffice. I propose :PersonInitiatesOffice and were HHDL to declare that he is the last of his line then :PersonTerminatesOffice

eroux commented 5 years ago

I'm trying to get a bigger picture in an RFC but I'm completely stuck I probably won't be able to move forward with it... I think the subclass of :Role (maybe :Incarnation instead of :TulkuLine) is the only necessary change for now (in order to be able to record that during the migration).

Then, during the migration, a possibility would be to make the use pattern of :personEventRole consistent between the various types of Roles, for instance we could add:

bdr:P197 :personEvent [ a :PersonEventAssumesOffice ;
          :personEventRole bdr:R8LS12876
          ] .

(which is currently absent and is inferred by the code through the :incarnationOf properties).