making quotationsbook.com zotero-aware

quotationsbook
Hi folks

I need your advice on building the facility to make our content Zotero-aware on quotationsbook.com. I've used sample data through the COinS generator with these key-value pairs:

Key Value
ctx_ver Z39.88-2004
rfr_id info:sid/ocoins.info:generator
rft_val_fmt info:ofi/fmt:kev:mtx:dc
rft.title Quotation by authorname
rft.creator authorname
rft.subject subjecttags of quote
rft.description quote text
rft.publisher http://quotationsbook.com
rft.type quotation
rft.language en
rft.rights Creative Commons

resulting in:

<span
class="Z3988"
title="ctx_ver=Z39.88-2004
&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Adc
&rfr_id=info%3Asid%2Focoins.info%3Agenerator
&rft.title=Quotation+by+authorname
&rft.creator=authorname
&rft.subject=subjecttags+of+quote
&rft.description=quote+text
&rft.publisher=http%3A%2F%2Fquotationsbook.com
&rft.type=quotation&rft.language=en
&rft.rights=Creative+Commons+">
<a><href="http://quotelink.com">quote text</a>
</span>

Am I missing anything here? All of these will of course be substituted with their real values. The idea is to put this markup on the quote display for over 43,000 items, this one for example:

http://quotationsbook.com/quote/29769

Regards
Amit
  • noksagt
    When you last asked this, I suggested you look into using unAPI for this. That is still my suggestion.

    Support for Dublin Core in COinS in Zotero exists, but it is somewhat rudimentary (but, in supporting DC at all, they are well ahead of the curve). COinS was not intended and may not be sufficient to be particularly rich metadata (with quotes, etc.). DC improves this a bit, but there are still relatively few producers and even fewer processors of DC-in-COinS & Zotero won't yet do what you want it to do with DC COinS.

    Your COinS could be improved too:

    rfr_id SHOULD NOT BE the ocoins.info generator unless you actually use that to generate your COinS. You should create an rfr_id for your program. (You're not the first to do this ;-)).

    "quotation" is neither a zotero type, nor has it been enumerated as a suggested DC type.
  • Tom Morris
    edited June 22, 2007
    This is an utterly bizarre way of doing it. Why doesn't Zotero simply support straight-forward RDF? For import and export of RDF, Zotero must have some kind RDF parser built in, so why not simply parse RDF from the web through eRDF/RDFa/GRDDL and DC metadata in the head? There's no way in hell anyone sane is going to sit down and add all this stuff, but something along the lines of:

    <span class="dc-creator">Arthur Schopenhauer</span>:
    "<span class="content">Compassion is the basis of morality.</span>"
    (<span class="dc-subject">morality</span>,
    <span class="dc-subject">compassion</span>)

    The classnames etc. are debatable, but the important thing is that it should be simple for anyone to do it. The current way is not simple, and it ought to be. It may be possible for the Library of Congress to use MODS XML or COinS (whatever *that* is!), but it's not simple for ordinary webmasters, whereas using classnames as microformats and eRDF does is a helluva lot simpler - which means people might actually do it.

    (Disclaimer: I have worked with Amit on the RDF release and I run the GetSemantic project)
  • noksagt
    edited June 22, 2007
    @Tom,

    Zotero does have some preliminary support for embedded RDF. The developers are also participating in the development of bibliographic ontologies and microformat.
    There's no way in hell anyone sane is going to sit down and add all this stuff, but something along the lines of:

    Arthur Schopenhauer:
    "Compassion is the basis of morality."
    (morality,
    compassion)

    The classnames etc. are debatable, but the important thing is that it should be simple for anyone to do it.
    And that debate is precisely why work is being done to make those formats standard. No such method exists that is in heavy usage. The ad hoc markup that does exist is not widely adopted & is often lacking in some way.
    COinS (whatever *that* is!)
    Context Objects in Spans is simple enough. The Wikipedia article and the OpenURL spec on this aren't THAT abstruse! There are generators for it, which means it is rather simple to add. COinS already has a reasonable level of adoption. I'd like to see wider adoption of UnAPI (it is also fairly straight forward (especially for webapp developers) & works right now.

    RDF is the path forward, but it isn't here yet.

This is an old discussion that has not been active in a long time. Instead of commenting here, you should start a new discussion. If you think the content of this discussion is still relevant, you can link to it from your new discussion.