How to write a post containing references to journal articles.
I’ve written an article recently about how to cite webpages in scientific articles. Here’s how to do it the other way around.
It’s easy enough to include a link to pubmed referencing the journal article you’d like to discuss(and put rev=’review’ in the link so it gets picked up by postgenomic), but we can get a lot fancier than that. You can actually embed citation metadata directly in your post, and at the same time, provide a link that will resolve to the article in the reader’s local library. This is especially handy for those journals that charge exorbitant rates for online access and for people using Zotero for citation management. Citing references like this is also required if you’re participating in the Bloggers for Peer-reviewed Research Reporting blog post aggregation system.
To embed the citation metadata, you need to put the info in a specially-formatted span tag. You don’t actually have to write this yourself, as there is a service that will take a PMID and return the correct tag. All you do is stick the PMID in the URL, like this, and it returns a code block you can stick in your post. Someone needs to write a plug-in for wordpress that takes PMIDs in span tags and converts them to the appropriate metadata.
Anyone visiting your page with a COinS activator extension will see a link that they can click on to find the article in their local library. To save the citation into Zotero, just click the folder or sheet icon that appears in the address bar of a COinS-containing page, next to where the RSS feed icon appears.
Those practicing open science can also use a WordPress plug-in to enable your own posts to be saved in this fashion. Here’s a screenshot of what it looks like, using my modification of Alf Eaton’s greasemonkey script.
How does this COinS generator work? I tried plugging in a PMID, but nothing happened. . .. what the deal yo?
G
If you use the form, it just tries to generate a tag based only on what you’ve entered. If you include the pmid in the HTTP GET request, like so: http://generator.ocoins.info/?id=pmid:17433565&sid=pubmed then it will resolve the PMID and spit out the tag with the corresponding info. In other words, you have to either pass the pmid in the URL or enter title, journal, page number, etc., in the form. It worked two weeks ago just by entering the PMID in the form, then they changed something.
I know it’s not very user-friendly, but I don’t think they expected it to see much use beyond librarians and other nerds, and here I go drawing attention to it.
Alf, in a comment on an earlier post, reminds me that Postgenomic uses rev=’review’ in the a element or the hReview microformat.
The hReview microformat is particularly good because is contains a means to distinguish between a post which is a review and individual reviews within the post. See the section on Object Includes. There’s no way to indicate this using COinS.
The problems I have with the hReview format is that they seem to want to use h2 elements for the summary, which works for short one line summaries but not for a paragraph summary, unless I were to restyle the h2 element to make it look like the body text style, but then that doesn’t make sense and would mess up anywhere else I would want to use it.
rev=’review’ is really simple, but should only be used when a post is about one item in particular. [EDIT: apparently I was wrong about this. Use it as much as you like.]
hReview has a mechanism for dealing with this, but if I’m going to be expected to remember a bunch of abbreviated class names I might as well embed RDF.
I’ve been a spectator for a while, and I’m just starting to actively use some of these technologies, but now that we’ve got some tools that can do useful things with metadata, such as Piggybank, Exhibit, Zotero, Postgenomic, etc the time is ripe for improvement on the metadata authoring side.
You didn’t mention the number one rule: cite the URL to the DOI, and not the URL on the publisher’s website. First of all, not all papers are on Pubmed; if they are, the PMID can be discovered based on the DOI…
You’re right about the DOI. It’s a far more universal identifier than the PMID.
What I’m trying to do is to show Joe Scienceblogger that there’s an easier and better way of identifying their review-containing posts than linking to a icon.
Using rev=’review’ in the link is a great and hidden-error(like mistyped page number) resistant way to do it, but I still feel like having the metadata in the post itself may be valuable, and since they’re all being set up to use COinS by the bpr3 people anyways, might as well explain how to do it right. Should I recommend putting the rev=’review’ in the DOI URL, or in the URL the DOI resolves to?
Of course, using the DOI instead of the PMID to generate the COinS as above works, too, but it lacks the article title and some additional fields that a PMID lookup has(I don’t know if that’s specific to that particular generator), so for the purpose of getting as much metadata as possible in the post itself, I’m recommending using the PMID, if available, to generate the data. This does result in the DOI being present in the COinS.
Pingback: Citing journal articles in blog posts and blog posts in journal articles at Synthesis
What do you think of OpenRef, invented and described here?:
http://baoilleach.blogspot.com/2008/01/doi-or-doh-proposal-for-restful-unique.html