Search and Cite for Science Bloggers

Crossref for WordPress

A couple of weeks ago I was reading a post by Will Griffiths on the ChemSpider Open Chemistry Web blog about how the DOI citation system of journal article lookups might be improved. The DOI system basically assigns each research paper a unique number depending, with an embedded publisher tag. Enter a DOI into a look up box (e.g. the DOI lookup on Sciencebase, foot of that page) and it almost instantaneously takes you to the paper in question. I use the DOI system for references in Sciencebase posts all the time.

There are a few cons that counteract its various pros, for instance, not all publishers use it and among those that do there are some who do not implement the DOI for their papers until they are in print, as opposed to online. Despite that it is very useful and commonly used. Having read Griffiths’ post about OpenURL a non-proprietary DOI alternative, I thought maybe it would be even more powerful if the concept were taken back another step the author level and I came up with the concept of a PaperID, which I blogged about on Chemspy.com. PaperID, I reasoned could be a unique identification tag for a reasearch paper, created by the author using a central open system (akin to the InChI code for labelling individual compounds). I’m still working out the ins and outs of this concept and while a few correspondents have spotted potentially fatal flaws others see it as a possible way forward.

Meanwhile, CrossRef, the association behind the publisher linking network, has just announced a beta version of a plugin for bloggers that can look up and insert DOI-enabled citations in a blog post. I’ve not investigated the plugin in detail yet, but you can download from a Crossref page at Sourceforge.net. the Crossref plugin apparently allows bloggers to add a widget to search CrossRef metadata using citations or partial citations. The results of the search, with multiple hits, are displayed and you then either click on a hit to go to the DOI, or click on an icon next to the hit to insert the citation into their blog entry. I presume they’re using plugin and widget in the accepted WordPress glossary sense of those words as the plugin is available only for WordPress users at the moment with a MoveableType port coming soon.

According to Geoffrey Bilder, CrossRef’s Director of Strategic Initiatives, “CrossRef is helping jumpstart the process of citing the formal literature from blogs. While there is a growing trend in scientific and academic blogging toward referring to formally published literature, until now there were few guidelines and very few tools to make that process easy.” Well that reference to a jumpstart sounds like marketing-speak to me, as Sciencebase and dozens of other science blogs have been using DOI for years.

Whether or not I will get around to installing what amounts to yet another WordPress plugin I haven’t decided. I may give it a go, but if it works as well as is promised you will hopefully not see the join. Meanwhile, let me have your thoughts on the usefulness of DOI and the potential of OpenURL and PaperID in the usual place below.

Share on FacebookTweet about this on TwitterShare on Google+Share on LinkedInEmail this to someonePrint this page

15 thoughts on “Search and Cite for Science Bloggers”

  1. Thanks for your thoughts Santy, this is certainly going to require a lot more thought than my initial post. Yes, you’re probably right, most publishers do honour the DOI system asap, but I’ve found that in the journals I use and cite there are two major publishers in the niche who do not. That could change at any time, of course, and maybe already have, I didn’t check this week!

    db

  2. Hi David,

    I think there are three pieces that are missing or you may want to consider further in your paperID proposal:

    – the work that Institutional Repositories are doing to standardise unique identification of scholarly objects and scholarly creators (most of them from the academic world)

    – the peer-review process’s peculiarities

    – OpenURL revolvers have not solved the “appropriate copy” dilemma totally, in part due to issues intrinsically associated to the use of namespaces of identifiers in network-based APIs

    Also, according CrossRef logs, most of the major publishers and medium-sized publishers, including Inderscience, do assign DOIs to their papers as soon as they go online (they do not wait until the printing stage) I m aware that CrossRef can takes minutes or hours but not more than a couple of days, to publish new DOIs. Most of the DOI lookup faults are due to publisher-side errors.

    BTW, The “widget-based” CrossRef Citation Plugin released to the public by Geoff (in the breaks of the ticTOCs meeting :-) ) is still being improved. CrossRef is still working fixing some bugs reported for initial users (e.g. DB access is timing out when locating DOIs in Lucene lookups, the “Changed server name” option (under “Options” > “CrossRef Citation Plugin”) seems to neither have effect nor persistence.)

    Santy

  3. A successful PaperID system would mean libraries do not have to face the absurdity of paying for object metadata without getting access to that object (would you pay for the words: ‘Harry Potter, JK Rowling’ without getting the book?). Good luck with it.

  4. You’re probably right Mitch…it may splice into the e-labbook concept better than the actual publication process, per Jean-Claude’s external wiki-key-timestamps for lab-book entries.

    Good move with ChemRank, I guess everyone realised it had been buried in spam. Using Connotea should pull it out of the cr*pflood nicely and give you the chance to build it into a great resource.

    db

  5. Unfortunately from experience, users don’t like having to recreate a paper database from scratch.

    I’m speaking really early into the game on this one, but I’m recoding the chemrank site I made last year to only use Connotea’s database. Since Connotea allows users to submit hard URLs or DOIs, piggybacking on Connotea’s own indexing system may be viable.

    Mitch

  6. @JCB I think there is a lot of potential in this little idea of PaperID, although as I’ve said elsewhere I’ve not quite worked out the ins and outs of how to get it started as a viable and realistic supplant to DOI/OpenURL that would be controlled at the author, as opposed to publisher, end of the publication process

    db

  7. baoilleach, I just installed and ran the plugin, so I had a better idea of what it can do. I tried a Science paper I wrote about a few weeks ago, and it found two ScienceDirect entries, so it doesn’t bode well…nor could it find the PNAS paper I’m writing about for spectroscopynow.com given author name…

    Another annoyance is where it slots into the WordPress GUI sidebar, I’d prefer it above the fold while I’m writing a post.

    db

  8. Unfortunately, it’s only as good as the data the journal publishers have deposited with CrossRef. It seems that even CrossRef don’t have the author list in some cases. See comments by Geoffrey and Chuck of CrossRef on my blog post.

  9. David,
    We’re always willing to experiment new ways of packaging and indexing experimental info. In terms of DOI and author generation, that is pretty much what I see Nature Precedings doing and we’ve used it a few times to do that. I assume a DOI refers to a specific version of a document (although Precedings does allow uploading of multiple versions I am not sure if you get a new DOI every time). For experiment pages, we still don’t have a “final version” for most of them – I am working with students in the next few weeks to get to that point. Of course if we discover errors down the road we will correct them and the wiki version tracking system will keep a record of that. We would like to get the associated raw data off our server as well though to get third-party time stamps. Maybe ChemSpider can help with some of that down the road (reaction monitoring spectra for example).

  10. David, this is a great find. It looks like the plugin uses a public Web API to pull the information, which could be an even bigger deal for developers wanting to build this kind of functionality into arbitrary Web applications.

  11. I guess there’s always the possibility to expand any ID tag to include any type of document, but it will grow exponentially if we add blog posts, essays, PhD theses etc, saying that, maybe there would be some way of using the DNA security tags you can get these days for labeling personal possessions.

    db

  12. I have extrapolated the thoughts to Open Notebook Science. Well there may be PaperID the title alone appears limited to some specific “document”, likely paper or electronic. For example, look at the “papers” on Molecules (http://www.mdpi.org/molecules/). Those “papers” certainly need the appropriate PaperID but now consider Open Theses and blogposts of specific merit which could also be identified. So, you could think of this being outside of just Papers. Maybe a ReferenceID (but haven’t search Google for collisions). I support the idea.

Comments are closed.