What do I have to do to get my site to work with Zotero?

[Alex]

I am not a web developer, but I'm in charge of a library-type website
which provides access to federal research and publications. I'm also
a faithful Zotero user with my own research.

Currently, Zotero does not work with our website. I'd like to know
what my web developer needs to do to add this functionality. I'm
thinking it may be a matter of changing the way we display search
results and citation information about the resources in our
collection.

Thanks for any help you can provide. Tech-savvy answers are fine, I
can pass them onto my developer if I don't understand them myself.

The site URL is www.ruralhealthresearch.org.

[Deborah Maron]

Hi Alex,

Thanks for your question.

There are three major options available to you,
1)Use a Zotero-ready web application
2)Expose your meta data using an open standard
3)Create a translator

The three options are further explained here:
Making your site Zotero ready: http://www.zotero.org/support/dev/make_your_site_zotero_ready

If you guys decide to take the translator route, here is some good documentation contributed by a community member about how to create translators that you can pass on to your developers:

Creating Translators: http://niche-canada.org/member-projects/zotero-guide/chapter1.html
 
and a page from us: http://www.zotero.org/support/dev/translator_overview

Hope this is helpful. Please let me know if you have any other questions.

Best,

Debbie

--
You received this message because you are subscribed to the Google Groups "zotero-dev" group.
To post to this group, send email to zoter...@googlegroups.com.
To unsubscribe from this group, send email to zotero-dev+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/zotero-dev?hl=en.

--
Deborah Maron, MA, MSLIS
Technology Outreach Specialist at George Mason University
4400 University Dr.
Fairfax, VA 22030

[Avram Lyon]

It would be very easy to Zotero-enable your site using COinS, but
unfortunately COinS doesn't support the Report type, which appears to
be the best type for most of your publications.

The best option then will probably by unAPI; Debbie's links have
information on how to set that up, and you can post here for advice as
well.

- Avram

2011/1/27 Deborah Maron <dma...@zotero.org>:

[Bruce D'Arcus]

On Thu, Jan 27, 2011 at 2:30 PM, Avram Lyon <ajl...@gmail.com> wrote:
> It would be very easy to Zotero-enable your site using COinS, but
> unfortunately COinS doesn't support the Report type, which appears to
> be the best type for most of your publications.
>
> The best option then will probably by unAPI; Debbie's links have
> information on how to set that up, and you can post here for advice as
> well.

Another plug for RDFa (though to the OP, I believe this won't work yet):

<p about="http://example.org/reports/1" typeof="bibo:Report">
<span property="dc:title">Report Title</span>,
<a rel="dc:publisher" href="http://example.org"><span
property="foaf:name">Example Organization</span></a>,
<span property="dc:date" content="2010-10-23">Oct 23, 1010</span>
</p>

But for publishers, a) it offers a lot more flexibility than COiNS,
and b) is easy(ish).

Bruce

[Avram Lyon]

2011/1/27 Bruce D'Arcus <bda...@gmail.com>:

> Another plug for RDFa (though to the OP, I believe this won't work yet):
>
> <p about="http://example.org/reports/1" typeof="bibo:Report">
>   <span property="dc:title">Report Title</span>,
>   <a rel="dc:publisher" href="http://example.org"><span
> property="foaf:name">Example Organization</span></a>,
>   <span property="dc:date" content="2010-10-23">Oct 23, 1010</span>
> </p>

I definitely agree that RDFa is a much nicer approach than COinS and
much easier to implement on small sites than unAPI.

But we need to write support for it. Would you be interested in taking
a crack at it, Bruce?

Avram

[Bruce D'Arcus]

I don't really have the time or the skill. But I can certainly help if
someone else has either or both.

And it seems like ideally we might want to build it on top of the
rdfa-api API code I posted yesterday? Am not sure where that stands
ATM.

Bruce

[Bruce D'Arcus]

Just a little followup on this ...

The API itself appears to be moving towards stability, if not
completely done, and the implementation is still missing some features
(like, I think, an actual RDFa syntax parser!).

But the basic conceptual approach you'd want to take here, it seems to
me, is to use the API implementation to do the parsing and grab the
objects, and then write some little adapter that maps that into
Zotero.

So, for example, with the api you'll be able to do ...

document.getElementsByType("bibo:Document")

That'll grab all descriptions of that type, and when there's inference
added (it's on the roadmap), it'll also grab all items that are
subclasses of bibo:Document.

Or, you could do:

document.getItemsByProperty("dc:title")

In either case, the objects you get back are, I believe, just a list
of properties that you can then map to Zotero's model, which is just a
mirror image of the logic that maps Zotero's model to BIBO RDF (the
RDF/XML and RDFa are the same model, just different syntax).

Bruce

[skornblith]

This API looks nice, and it might eventually be useful for someone to
port our existing RDF support over to it, but it doesn't seem to do
much for us, since all we need is the RDFa syntax parser as well. We
already have a Bibliontology translator that is capable of importing
data that's loaded into Zotero's Tabulator-backed triple store.
Getting the triples out of HTML/RDFa is the part that someone needs to
implement.

Simon

On Jan 27, 5:32 pm, "Bruce D'Arcus" <bdar...@gmail.com> wrote:
> Just a little followup on this ...
>
>
>
>
>
>
>
>
>
> On Thu, Jan 27, 2011 at 3:40 PM, Bruce D'Arcus <bdar...@gmail.com> wrote:
> > On Thu, Jan 27, 2011 at 3:18 PM, Avram Lyon <ajl...@gmail.com> wrote:

> >> 2011/1/27 Bruce D'Arcus <bdar...@gmail.com>:

[Bruce D'Arcus]

On Thu, Jan 27, 2011 at 11:07 PM, skornblith <si...@simonster.com> wrote:
> This API looks nice, and it might eventually be useful for someone to
> port our existing RDF support over to it, but it doesn't seem to do
> much for us, since all we need is the RDFa syntax parser as well. We
> already have a Bibliontology translator that is capable of importing
> data that's loaded into Zotero's Tabulator-backed triple store.
> Getting the triples out of HTML/RDFa is the part that someone needs to
> implement.

I'm confused. You say:

You have an RDFa parser, which IS "getting the triples out ..."

You also have a bibo translator.

So exactly what's missing?

Bruce

[skornblith]

We have an RDF parser and a BIBO translator. We don't have an RDFa
parser.

[Alex McEllistrem-Evenson]

Thank you, everyone, for your assistance.

After doing some more research and consulting with my developer, we have determined that the best way to proceed is to include all of our citation information in our metadata and make it available for Zotero to read.

Can any of you send me a URL of a site that operates this way, so we can look to see how they structure their metadata?

Thank you in advance.

Alex

[Ed Summers]

On Fri, Jan 28, 2011 at 2:05 PM, skornblith <si...@simonster.com> wrote:
> We have an RDF parser and a BIBO translator. We don't have an RDFa
> parser.

Not sure if jQuery is already a dependency in Zotero, but I've heard
good things about:

http://code.google.com/p/rdfquery/

//Ed

[Bruce D'Arcus]

+1

So once we have that, it should be a pretty short jump; right?

Bruce

> //Ed

[Richard Karnesky]

> After doing some more research and consulting with my developer, we have
> determined that the best way to proceed is to include all of our citation
> information in our metadata and make it available for Zotero to read.

There is still the question of the format you should expose your
metadata in, which this thread has been discussing. There are most
likely good sites that demonstrate using RDF, but I'm not personally
aware of any (does anyone elsehave good examples?).

The application I co-develop, refbase, uses UnAPI and MODS XML to
expose nearly all metadata to Zotero. It seems robust & our UnAPI and
MODS XML components are stand-alone enough that each has been used in
other GPL web apps. We do plan to serve RDF over UnAPI in the future
& will have embedded RDF in the more distant future.

--Rick

[Bruce D'Arcus]

I don't have a ton of experience with it, would think the preferred
way to publish non-RDFa RDF (RDF/XML, turtle, etc.) is via
content-negotiation, to avoid layering on another API.

E.g. browser requests a page for http://example.org/123, and it gets
an HTML representation (123.html), an RDF consumer instead gets an
RDF/XML representation (123.rdf), etc. (123.json). That's how my site
works in any case*.

Maybe Ed has something to add here?

My reason for suggesting RDFa is that it seems to give the advantages
of RDF (namely, flexibility) without the hassles of serving up
separate files, and using content-negotiation.

Bruce

* http://bruce.darcus.name

[skornblith]

On Jan 28, 4:19 pm, "Bruce D'Arcus" <bdar...@gmail.com> wrote:
> On Fri, Jan 28, 2011 at 2:18 PM, Ed Summers <e...@pobox.com> wrote:
> > On Fri, Jan 28, 2011 at 2:05 PM, skornblith <si...@simonster.com> wrote:
> >> We have an RDF parser and a BIBO translator. We don't have an RDFa
> >> parser.
>
> > Not sure if jQuery is already a dependency in Zotero, but I've heard
> > good things about:
>
> >    http://code.google.com/p/rdfquery/
>
> +1
>
> So once we have that, it should be a pretty short jump; right?

Since we are using the Tabulator data store, we would need to adapt it
to that first, or alternatively, we'd need to adapt our code to
another RDF data store that can serialize to RDF/XML. It seems like
Tim Berners-Lee is already working on the former:

http://dig.csail.mit.edu/hg/tabulator/file/ccc61abc1cc1/chrome/content/js/rdf/rdfa.js

Unfortunately, from the comments at the top, it looks like it doesn't
work yet.

Simon