COiNs?

[clairestewart]

Hi, I've trolled through discussion archives but can't tell whether or
not COiNs is fully implemented for the Journal? Having a little
trouble reliably getting it for articles.

[Jodi Schneider]

Hi Claire,

Editors put COinS in on an article-by-article basis, so a few things could slip through. We also general don't add COinS for open access articles and web pages.

That said--could you give an example of the article(s) where it's not working?

-Jodi

[clairestewart]

Thanks Jodi,

That makes sense since I was initially looking at older content ...I
was starting with an article from issue 2 (Geocoding LCSH ... <http://
journal.code4lib.org/articles/52>) and on my first shot I only
extracted the "Impact factors..." article rather than the md for the
article itself. Tried a different technique and I'm getting it into
Zotero successfully now; though some of the data is not as I'd expect
it's an easy fix for me.

Thanks again
Claire

On Nov 8, 11:45 am, Jodi Schneider <jodi.a.schnei...@gmail.com> wrote:
> Hi Claire,
>
> Editors put COinS in on an article-by-article basis, so a few things could
> slip through. We also general don't add COinS for open access articles and
> web pages.
>
> That said--could you give an example of the article(s) where it's not
> working?
>
> -Jodi
>
> On Sun, Nov 8, 2009 at 3:43 PM, clairestewart <
>

[Tom Keays]

I don't know enough about how Zotero handles COinS to really give a
good analysis. I can say that the COinS is this article was
constructed using the generator program, http://generator.ocoins.info/
and it is well-formed and rich in metadata for the article.

However, COinS spans aren't themselves designed to work with citation
management programs since the metadata fields are are all optional.
Zotero does better, I think, with the DOI that was also supplied in
the article since CrossRef now provides good underlying metadata for
creating citations and Zotero presumably uses that to populate the
appropriate citation fields. That said, the COinS span from Chris's
article does include the DOI, so Zotero ought to digest it fine.

Tom

[Jodi Schneider]

Claire,

Ah! That makes sense. First I thought you were talking about references in the bibliography. I think the COinS for each article comes from unAPI:
http://lackoftalent.org/michael/blog/unapi-wordpress-plug-in/
based on our list of plugins:
http://wiki.code4lib.org/index.php/Code4Lib_Journal_WordPress_Customizations

If you notice consistent problems, we may be able to tweak the configuration, or the plugin author Mike Giarlo (a #code4lib regular) might be able to help.

-Jodi

[Tom Keays]

The only COinS <span class="Z3988"> tag that I could find in the
article in question is the reference to a Nature article in Notes
section which was inserted by hand by the editor (me).

We are using the unAPI Wordpress plugin that Jodi references; Zotero
also notes it in their dev pages.

http://www.zotero.org/support/dev/wordpress

By itself, all the unAPI plugin does is generate a link tag in the
header referencing the server,

<link rel="unapi-server" type="application/xml" title="unAPI"
href="http://journal.code4lib.org/wp/wp-content/plugins/unapi/server.php"/>

and a bit of metadata in the body of the article, e.g.,

<abbr class="unapi-id"
title="http://journal.code4lib.org/articles/52"><!-- &nbsp; --></abbr>

If I'm reading this right, for Zotero to really use unAPI, in the
context of creating a COinS for a blog post, there needs to be another
"COinS button" plugin installed. And, looking at the plugins folder in
our installation, I'm not seeing that.

Tom

[Jonathan Brinley]

We don't actually have COinS for each article. We just have it
wherever the editor decides to insert it.

unAPI is a different matter entirely. The plugin adds the <link> and
<abbr> tags that let, for example, Zotero know that it can go to
http://journal.code4lib.org/wp/wp-content/plugins/unapi/server.php?id=http://journal.code4lib.org/articles/52&format=oai_dc
to get data about the article.

Last I checked, anyway, Zotero would only look at the unAPI data and
not the COinS data. The latter is for people with browser plugins
(e.g., http://libx.org/) that can form an appropriate link from the
COinS <span> tag, not for Zotero.

Have a nice day,
Jonathan

> --~--~---------~--~----~------------~-------~--~----~
> You received this message because you are subscribed to the Google Groups "Code4Lib Journal-discuss" group.
> To post to this group, send email to c4lj-d...@googlegroups.com
> To unsubscribe from this group, send email to c4lj-discuss...@googlegroups.com
> For more options, visit this group at http://groups.google.com/group/c4lj-discuss?hl=en
> -~----------~----~----~----~------~----~------~--~---
>
>



--
Jonathan M. Brinley

jonatha...@gmail.com
http://xplus3.net/

[Emily Lynema]

Zotero can easily import COiNs data, but it might choose unAPI over
COiNS, if unAPI is enabled. We use COiNs tags in our catalog to let
people import references into Zotero.

-emily

Emily Lynema
Associate Department Head
Information Technology, NCSU Libraries
919-513-8031
emily_...@ncsu.edu

[Tom Keays]

Thanks for that explanation. It's nice to know how that bit works.

Tom

> --

>
> You received this message because you are subscribed to the Google Groups "Code4Lib Journal-discuss" group.

> To post to this group, send email to c4lj-d...@googlegroups.com.
> To unsubscribe from this group, send email to c4lj-discuss...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/c4lj-discuss?hl=.
>
>
>

[Jonathan Rochkind]

Yeah, the point of the COinS isn't to support citation extraction, our
purpose in adding COinS is to support direct links to link resolvers to
get locally licensed full text.

So we only add COinS for articles that are not publically accessible.
It's a bit of work to add structured metadata for citations, so we
prioritize those that meet the actual original purpose of the COinS
standard.

In my opinion, COinS isn't a very good solution for extracting citations
anyway; the COinS (OpenURL key-encoded-value) standard doesn't
neccesarily allow you to add all the data elements you'd want in an
extracted citation anyway.

If you want to extract the citation for the Code4Lib Journal article
itself (as opposed to references cited), we are including UnAPI support
for that purpose, which should be picked up by Zotero. If that's not
working, perhaps our system has a flaw.

But our goal is UnAPI for the C4LJ articles themselves; COinS for
references cited that are not public access; nothing for references
cited that are public access.

Jonathan

[Jodi Schneider]

Right, but sometimes one wants to cite the Code4Lib Journal article itself.

That's what the unAPI does. The test I just did pulls in title, abstract, and date well. Meanwhile authors are ok but don't get parsed individually. The thing most in need of fixing is that it parses as a Web Page (with The Code4Lib Journal as the 'website title'). Anybody want to take that on?

-Jodi

--

You received this message because you are subscribed to the Google Groups "Code4Lib Journal-discuss" group.

To post to this group, send email to c4lj-d...@googlegroups.com.
To unsubscribe from this group, send email to c4lj-discuss...@googlegroups.com.

For more options, visit this group at http://groups.google.com/group/c4lj-discuss?hl=.