GeoscienceAustralia / Placenames-Ontology

Codes and other documentation for transformation, querying and managing the Place Names dataset in Linked Data format.
Creative Commons Attribution 4.0 International
2 stars 3 forks source link

Document alignment of PlaceNames ontology to ISO 19112 abstract model #27

Open dr-shorthair opened 4 years ago

dr-shorthair commented 4 years ago

ISO 19112 defines a standard model for 'geographic identifiers', such as place-names. The PlaceNames ontology is an implementation of a profile. The correspondence to the ISO 19112 model should be explained.

image

ibastrak commented 4 years ago

I'm not familiar with this one. Is this diagram for the 2019 version? Should we suggest it for adoption to IT004 if not adopted yet. I am thinking about connecting 'language' to it.

dr-shorthair commented 4 years ago

No - this is from the old 2003 version. I don't have a copy of 2019.

dr-shorthair commented 4 years ago

I found this diagram from 2nd edition here http://iso.sparxcloud.com/index.php

image

AFAICT the only substantive changes are

Note that the management of the actual 'name' value depends on GeographicIdentifier , and in turn on the MD_Identifier from 19115-1

image

ibastrak commented 4 years ago

Interesting. Yes, having GeographicIdentifier in Location multiple times is very helpful. it also provides multiple options for locale

dr-shorthair commented 4 years ago

The approximate correspondence of ISO 19112 with the PlaceNames ontology is

iso:Location == pn:Place
iso:GeographicIdentifier == pn:PlaceName
iso:MD_Identifier ~ rdf:langString

though some of the other attributes of iso:MD_Identifier we would attach to `pn:Placename instead.

In the PlaceNames ontology the cardinality of hasPlaceName is unspecified. Which makes it [0..*]. That allows for multiple names, as shown in http://linked.data.gov.au/def/placenames#Place. But also for no name. Are there places without names? Of course, but we probably aren't interested in them, at least not in the context of a gazetteer. So maybe that cardinality should be tweaked to [1..*].

ibastrak commented 4 years ago

There are many unnamed places which are still places of interests. They will be unique features in combination with location. Preps, we can ignore them for Gazetteer.

dr-shorthair commented 4 years ago

The key characteristics of pn:Place are

  1. its classification(s) (i.e. type)
  2. its name(s)

if it is missing either of those, then it probably is not a 'Place' in the context of a Gazetteer? It is just a Feature.