Group: http://groups.google.com/group/pelagios-project/topics
Simon Rainer <Rainer...@ait.ac.at> Oct 14 06:45AM
--
Dear list,
just a small reminder for anyone (especially those who have taken part in our Gazetteer Workshop at ISAW in September :-) to check the draft of our "Gazetteer Interconnection Format". Any feedback would be greatly appreciated!
https://github.com/pelagios/pelagios-cookbook/wiki/Pelagios-Gazetteer-Interconnection-Format
Does this meet your understanding of what we discussed in New York? Do you see any errors/issues/gaps?
Question specifically for the gazetteer partners: will is this work for you as a format to export your data? Is the Wiki page sufficient information you need to get started?
Cheers,
Rainer
RAINER SIMON
Scientist
Safety and Security Department
Information Management
AIT Austrian Institute of Technology GmbH
Donau-City-Strasse 1 | 1220 Vienna | Austria
T +43 664 2351792 | F +43 50550-4150
rainer...@ait.ac.at<mailto:rainer...@ait.ac.at> | http://www.ait.ac.at<http://www.ait.ac.at/>
FN: 115980 i HG Wien | UID: ATU14703506
http://www.ait.ac.at/Email-Disclaimer
You received this message because you are subscribed to the Google Groups "pelagios" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pelagios-proje...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
Hi Lex, Yes, we'll definitely want to distinguish between different gazetteers as you say. Hm - the handle.net issue is a really good point. I'll also need to think about this. But I guess there needs to be at least some sort of base URI (defined in the dump file) that is unique to each gazetteer. We can then use this as "context", and handle each gazetteer as a "named graph" (to use semantic Web speak). After that, it's no longer a problem if different gazetteers use the same URI prefixes.
<And I am assuming that evidence, annotations, toponym characteristics, language and other info about each attestation of ancient places is held in the canonical gazetteer records> Yes, this doesn't need to be included, especially when it comes to things such as evidence, details about attestations etc. From the Pelagios perspective, we (i.e. Leif, Elton, Pau & myself) consider everything beyond the most basic gazetteer metadata beyond our expertise, as it were; and I don't think it makes sense for us to force too much of a detailed schema on anyone (nor would we be in a position to do it.) That being said: the nice thing about RDF is of course that you can always add additional stuff to your data dumps without breaking anything. So if there's anything you think makes sense to include for the wider Linked Data community (even if Pelagios might not immediately make use of it) you can always do that.
Tom's point about language codes is a good example for this: the @ syntax is in fact "baked into" RDF. We'll be treating that as optional in Pelagios (although our parser picks it up, and I'm pretty sure we'll make use of it if its there). You can use the full list of ISO 639-2 codes. In other words: <http://www.my-gazetteer.org/places/12345678/names/0001> skos:label "平海" . <http://www.my-gazetteer.org/places/12345678/names/0001> skos:label "平海"@zh . are both valid place names for us. Cheers, Rainer
Hi Lex,
< I was just wondering what goes into the "description" element in my case... >
In the end, it’s very much up to you what you want to put in there. But I think you’re hitting the nail on the head with regard to disambiguation. It should be something that allows human users to instantly disambiguate/confirm. (E.g. example from Pleiades: Thebes – “The ancient city of Thebes in Boeotia (modern Greece)”.) So your suggestion with parent jurisdiction sounds indeed a good choice for CHGIS!
< In our case we have these complications, like Tibetan names transcribed into Chinese >
Hm – I see. Another very good point. I think this is very relevant for other gazetteers, too. Not sure how to approach it off the top of my head. Would it makes sense if we devised some sort of extended syntax, so we can express cases of “name in language A, transcribed/transliterated to language B”?
Cheers,
Rainer