Groups keyboard shortcuts have been updated
Dismiss
See shortcuts

Re: [pelagios] Digest for pelagios-project@googlegroups.com - 1 Message in 1 Topic

4 views
Skip to first unread message

Merrick Lex Berman

unread,
Oct 15, 2013, 10:05:16 AM10/15/13
to pelagios...@googlegroups.com, Simon Rainer
Hi Simon,

Thanks for sending!   I have to think about this for a week or so before replying, but I had a question in mind in recent days:   Shouldn't there be a namespace declaration of some kind?    Maybe we need an explicit way to search only on one of the gazetteers registered in Pelagios, such as Plieades, PastPlace, or CHGIS, etc.   And there may be other gazetteer bases in the future...    URIs might not be enough, imagine if two projects are using hdl.handle.net URIs, there would be no way to distinguish the source from the URI base itself.

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, so that doesn't need to be included?    Or is the idea that Tom Elliott raised about being able to explicitly record the language of the toponym going to be part of the schema?     I think he showed an example using an @ sign, like:   平海@zh   or some syntax that declared the language.  

Thanks,
--Lex
   
On 10/14/2013 11:04 AM, pelagios...@googlegroups.com wrote:

Group: http://groups.google.com/group/pelagios-project/topics

--
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.


--
ॐ - Lex Berman
desk (617) 496-9439 - info
Web Services Manager, Center for Geographic Analysis

watershed: Alewife Brook > Mystic River > Boston Harbor > Atlantic Ocean

Simon Rainer

unread,
Oct 15, 2013, 10:51:34 AM10/15/13
to Merrick Lex Berman, pelagios...@googlegroups.com
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





________________________________________
Von: Merrick Lex Berman [mbe...@fas.harvard.edu]
Gesendet: Dienstag, 15. Oktober 2013 16:06
An: pelagios...@googlegroups.com
Cc: Simon Rainer
Betreff: Re: [pelagios] Digest for pelagios...@googlegroups.com - 1 Message in 1 Topic

Hi Simon,

Thanks for sending! I have to think about this for a week or so before replying, but I had a question in mind in recent days: Shouldn't there be a namespace declaration of some kind? Maybe we need an explicit way to search only on one of the gazetteers registered in Pelagios, such as Plieades, PastPlace, or CHGIS, etc. And there may be other gazetteer bases in the future... URIs might not be enough, imagine if two projects are using hdl.handle.net URIs, there would be no way to distinguish the source from the URI base itself.

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, so that doesn't need to be included? Or is the idea that Tom Elliott raised about being able to explicitly record the language of the toponym going to be part of the schema? I think he showed an example using an @ sign, like: 平海@zh or some syntax that declared the language.

Thanks,
--Lex

* Pelagios gazetteer "interconnection format" - feedback requested [1 Update]

Pelagios gazetteer "interconnection format" - feedback requested<http://groups.google.com/group/pelagios-project/t/70185a1b907e331f>

Simon Rainer <Rainer...@ait.ac.at><mailto: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><mailto: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<mailto:pelagios-proje...@googlegroups.com>.
For more options, visit https://groups.google.com/groups/opt_out.


--
ॐ - Lex Berman
desk (617) 496-9439 - info<http://www.iq.harvard.edu/people/lex-berman>

Merrick Lex Berman

unread,
Oct 15, 2013, 11:19:35 AM10/15/13
to Simon Rainer, pelagios...@googlegroups.com
On 10/15/2013 10:51 AM, Simon Rainer wrote:
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.

Okay, as long as we have a way to filter them, it should be fine.

<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.

I agree about the simplicity for Pelagios.   I was just wondering what goes into the "description" element in my case...   Possibly, the parent hierachy?   like: 

"Historical administration  Xiushan Ju >> Dali Guo.   Present Location:   Tonghai Xian >> Yuxi Diqu >> Yunnan Sheng >> China [P.R.C.]"

not sure what is most useful...  

Note:  I think parent jurisdiction is one of the most useful ways to disambiguate placenames.   So I guess that would be the best thing to include in the Description...?


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

Sounds good.  I hope to hear more from Tom about how he uses that.   In our case we have these complications, like Tibetan names transcribed into Chinese,  "Qamdo"    so is it:  "昌都"@bo    or:   "昌都"@zh  ?   Similarly, Thai names, like:   Sipsongpanna...   "西双版纳"@th    or    "西双版纳"@zh   

Anyway, this is not important for now, but I want to get the usage right...

best,
--Lex


    


Simon Rainer

unread,
Oct 16, 2013, 2:19:35 AM10/16/13
to Merrick Lex Berman, pelagios...@googlegroups.com

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

Reply all
Reply to author
Forward
0 new messages