Populate geo-referenced places

[kevinstadler]

The targets of "place corresponds to" assertions (IRIs from Pleiades and Geonames) are currently lose nodes in the triple store without an explicit CRM class. Two things are needed to make them usable for visualisatons:

• since they currently have no CRM class they do not match the strict typed query (see query link. This is a similar situation to the social relationships in Fix social relationship model query #78. Two possible fixes: either mark the IRIs with the CRM class (Site), OR relax the queries (following which pattern?) to not check for the model classes
• 242 locations currently have coordinates (in lat,long format) attached to them through spatiotemporal existence assertions generated by Lukas: link

[tla]

For #78 it was a clear bug and the class statements should have been there. For this case it is a lot less clear-cut - we are asserting that our E27 Site (which is an IRI with that class designation) corresponds to some Pleiades IRI, but that doesn't necessarily make the other thing a Site; it might be considered by Pleiades to merely be an E53 Place, for example.

[tla]

I have changed these in the pathbuilder to have an E1 Entity as the expected object type.