guidelines for publishers
ed.su...@gmail.com
metadata to Zotero in their newly minted Solr powered catalog. Is the
exposing_your metadata page comprehensive [2]? It seems to be lacking
some details, particularly because it buries COinS support inside
blogging software, and doesn't mention unAPI at all. I seem to
remember there other standard ways folks can expose their metadata so
zotero can grab it?
I would recommend you create a "For Publishers" tab in your menu right
next to "For Developers" and detail all the standard mechanisms a
publisher could use to expose their metadata to enable Zotero to
recognize it. Or something like that :-)
//Ed
[1] http://siris-collections.si.edu/search/
[2] http://dev.zotero.org/exposing_your_metadata
Sean Gillies
currently used to list some applications that expose their metadata.
Speaking of: I don't understand why COinS is recommended. Instead of
embedding a link to a metadata web service in your web page, why not
just embed the metadata itself as RDF? Works great for feed discovery in
blogs.
Cheers,
Sean
Sean Takats
frequent requests for the same kinds of information about how to
tweak OPAC software, etc. to improve Zotero readability. A
publishers' section would save us a lot of time spent responding to
these questions, and the wiki model would allow for refinement and
community contributions.
I also definitely agree that embedded RDF is the way to go. One
current problem is that we don't yet have an embedded RDF model that
preserves all of a Zotero item's data. It seems to me that we want
something that allows for lossless roundtrips between import and
export from Zotero as embedded RDF. The stumbling block right now is
the degree to which such an embedded RDF draws on existing schema.
We could more quickly move toward a more unilateral solution to
replicate the entire Z model (e.g. just embed our current RDF export
in web pages and support the detection and import of those embedded
items), but there is also value in promoting a more "open" model
which draws on existing formats.
Sean
Sean Gillies
of the COinS format, but it does embed the metadata right in the document.
Cheers,
Sean
Bruce D'Arcus
...
> I also definitely agree that embedded RDF is the way to go. One
> current problem is that we don't yet have an embedded RDF model that
> preserves all of a Zotero item's data. It seems to me that we want
> something that allows for lossless roundtrips between import and
> export from Zotero as embedded RDF. The stumbling block right now is
> the degree to which such an embedded RDF draws on existing schema.
> We could more quickly move toward a more unilateral solution to
> replicate the entire Z model (e.g. just embed our current RDF export
> in web pages and support the detection and import of those embedded
> items), but there is also value in promoting a more "open" model
> which draws on existing formats.
We're closing in on a first draft of the biblio ontology which is
intended to work for this.
Now would be a good time for feedback.
Alas, it might be a little funky (in particular verbose for, say,
contributors) to do this in eRDF instead of linked RDF/XML.
BTW, I'll be extracting a snapshot of this soon for inclusion in
OpenDocument.
Bruce
Bruce D'Arcus
some people working on it).
This XHTML:
<p about="http://ex.net/1">
<span rel="bibo:contribution">
<a rel="bibo:contributor" href="http://ex.net/people/1">
<span property="foaf:name">Doe, J.</span>
</a>
<a rel="bibo:role" href="http://ex.net.roles/author"/>
</span>
<span property="dc:date">2000</span>
<span property="dc:title">Some Title</span>
<a rel="dc:isPartOf" href="http://nytimes.com">
<span property="dc:title">New York Times</span>
</a>
</p>
Gets extracted as this RDF/XML:
<rdf:RDF
xmlns:foaf='http://xmlns.com/foaf/0.1/'
xmlns:rdf='http://www.w3.org/1999/02/22-rdf-syntax-ns#'
xmlns:dc='http://purl.org/dc/elements/1.1/'
xmlns:bibo='http://purl.org/ontology/bibo/'>
<rdf:Description rdf:about="http://ex.net/1">
<dc:date>2000</dc:date>
<bibo:contribution>
<rdf:Description>
<bibo:contributor rdf:resource="http://ex.net/people/1"/>
<bibo:role rdf:resource="http://ex.net.roles/author"/>
</rdf:Description>
</bibo:contribution>
<dc:isPartOf>
<rdf:Description rdf:about="http://nytimes.com">
<dc:title>New York Times</dc:title>
</rdf:Description>
</dc:isPartOf>
<dc:title>Some Title</dc:title>
</rdf:Description>
<rdf:Description rdf:about="http://ex.net/people/1">
<foaf:name>Doe, J.</foaf:name>
</rdf:Description>
</rdf:RDF>
The contribution encoding is more complex than I'd like, but I don't see
any way around it.
Would be interesting to compare with eRDF.
Bruce
bill mckinney
What would be wrong with using simple, plain old meta tags in the head
section of html? Meta tags have been around forever and are way easy
to extract in a Zotero translator. Here is a really beautiful example
from The New England Journal of Medicine (http://content.nejm.org/cgi/
content/short/357/8/753):
<meta name="citation_journal_title" content="The New England Journal
of Medicine">
<meta name="citation_authors" content="Adams, Ted D.; Gress, Richard
E.; Smith, Sherman C.; Halverson, R. Chad; Simper, Steven C.;
Rosamond, Wayne D.; LaMonte, Michael J.; Stroup, Antoinette M.; Hunt,
Steven C.">
<meta name="citation_title" content="Long-Term Mortality after Gastric
Bypass Surgery">
<meta name="citation_date" content="08/23/2007">
<meta name="citation_volume" content="357">
<meta name="citation_issue" content="8">
<meta name="citation_firstpage" content="753">
<meta name="citation_doi" content="10.1056/NEJMoa066603">
<meta name="citation_pdf_url" content="http://content.nejm.org/cgi/
reprint/357/8/753.pdf">
<meta name="citation_abstract_html_url" content="http://
content.nejm.org/cgi/content/abstract/357/8/753">
<meta name="citation_fulltext_html_url" content="http://
content.nejm.org/cgi/content/full/357/8/753">
<meta name="citation_pmid" content="17715409">
<meta name="dc.Contributor" content="Adams, Ted D.">
<meta name="dc.Contributor" content="Gress, Richard E.">
<meta name="dc.Contributor" content="Smith, Sherman C.">
<meta name="dc.Contributor" content="Halverson, R. Chad">
<meta name="dc.Contributor" content="Simper, Steven C.">
<meta name="dc.Contributor" content="Rosamond, Wayne D.">
<meta name="dc.Contributor" content="LaMonte, Michael J.">
<meta name="dc.Contributor" content="Stroup, Antoinette M.">
<meta name="dc.Contributor" content="Hunt, Steven C.">
<meta name="dc.Title" content="Long-Term Mortality after Gastric
Bypass Surgery">
<meta name="dc.Identifier" content="10.1056/NEJMoa066603">
<meta name="dc.Date" content="08/23/2007">
<meta name="citation_publisher" content="Mass Med Soc">
Bruce D'Arcus
> No offense intended to RDF fans (I'm one to some degree), but...
>
> What would be wrong with using simple, plain old meta tags in the head
> section of html?
The reason I was experimenting with RDFa -- and why I believe the
approach superior to using meta tags in the head -- is because:
1) it's like micro-formats in the sense that it encodes visible content
as metadata (no duplication)
2) it's fully equivalent from a modeling perspective with the RDF
ontology we're working on (IIRC I was just cc-ing here from that list)
Bruce
bill mckinney
ontologies.
However, don't overlook the "for publishers" part of the thread title.
At the end of the day a publisher's electronic production department
has to buy into the ontology you eventually develop. A bunch of large,
corporate, well funded publishers may be able to pull off what you are
proposing. Many more smaller, privately held (but just as important)
publishers won't have the staff or budget to deliver embedded RDFa.
Believe me, I've been working for these types of publishers for 15
years. Consider getting early feedback on the RDF ontology approach
from a variety of publishers before you get too far.
Most any publishers can add meta tags to html today.
Bruce D'Arcus
> I understand the technical/intellectual interest and value in RDF
> ontologies.
It's thoroughly practical.
> However, don't overlook the "for publishers" part of the thread title.
> At the end of the day a publisher's electronic production department
> has to buy into the ontology you eventually develop. A bunch of large,
> corporate, well funded publishers may be able to pull off what you are
> proposing. Many more smaller, privately held (but just as important)
> publishers won't have the staff or budget to deliver embedded RDFa.
> Believe me, I've been working for these types of publishers for 15
> years. Consider getting early feedback on the RDF ontology approach
> from a variety of publishers before you get too far.
Anyone is free to join the effort, and it is explicitly aimed at some
practical balance of simplicity and expressiveness suitable for
publishers and consumers.
Beyond that, I'm not going to go out of my way to contact publishers
individually. I have no illusions that even if I did that every possible
player in this space would standardize on any one approach.
> Most any publishers can add meta tags to html today.
Sure, so they should do that; better than nothing. The DCMI has just
been working on updating their suggestions on embedding DC descriptions
in HTML. That ought to be able to cover most of the baseline stuff.
Bruce