Serving shortDOIs

The shortDOI service was launched by the International DOI Foundation (IDF) in May 2010. The service creates short versions of the often long DOIs, e.g. 10/dvq instead of 10.1093/hmg/ddp202 – written as URL this would be http://doi.org/dvq instead of http://dx.doi.org/10.1093/hmg/ddp202. shortDOIs started as a CrossRef Labs project in 2009 and were originally named TOI DOI – TOI stands for tiny object identifier. For a good introduction I recommend Eric Hellman’s blog post written in May 2010 when the shortDOI service was launched.

The long and the short of it

shortDOIs provide exactly the same service as normal DOIs, i.e. they redirect the user to the digital resource. They are convenient, particularly for email and Twitter where space is limited. Links to journal articles created by URL shorteners such as bit.ly or goo.gl look similar, but require two redirects (first to dx.doi.org, then to the journal article). But URL shorteners provide additional services to users, e.g. customized links and usage statistics.

As far as I can tell shortDOIs have not become popular since the service started more than a year ago. One important reason is certainly that publishers are not really using them for their journal articles. I don’t think many users will go through the extra steps creating a shortDOI just to use a DOI with Twitter – the Twitter URL shortener t.co will do this automatically for them. Michael Kuhn has created a bookmarklet that makes it a little bit easier to create shortDOIs. If we want shortDOIs to ever become popular, then we should ask journal publishers, bibliographic databases, reference managers and other places that currently use DOIs to enable them.

Starting yesterday ScienceCard is using shortDOIs instead of DOIs, and is also using the shortDOI to link to individual articles on ScienceCard, e.g. http://sciencecard.org/articles/dvq. Yesterday I’ve created about 750 shortDOIs for ScienceCard, and  this probably already makes ScienceCard one of the larger shortDOI users. I will decide in the coming months whether shortDOIs have improved the ScienceCard service. A nice feature would for example be the announcement of newly published papers via Twitter.

Related Posts Plugin for WordPress, Blogger...
This entry was posted in Conferences, Interviews, Presentations, Recipes, ResearchBlogging, Reviews, Snippets, Thoughts and tagged , . Bookmark the permalink.

8 Responses to Serving shortDOIs

  1. Bastian says:

    Maybe it’s a stupid question: But where is the benefit of using “normal” DOIs over ShortDOIs?

  2. Bastian, one advantage of “normal” DOIs is that everybody is using them, and this makes it easy to connect different systems together. This could of course change if more people use shortDOIs. Another difference is the prefix used by “normal” DOIs, e.g. 10.1371 for PLoS. But I personally don’t see this as an important feature. You can get this information and more by looking at the metadata for the DOI.

  3. Bastian says:

    Thanks! So the network effect may be the main reason to use the normal ones and not systemic features. Good luck in spreading the short version :)

  4. Bastian, you can in theory extract information out of DOIs, e.g. different versions of a paper. But because their is no standard for these things and every publisher can create DOIs to their liking (after the 10.x/ prefix), I don’t see this as an important feature.

    Something that would be good and that I haven’t seen yet is a shortDOI -> DOI service so that you can translate to the DOI for the services that require it.

  5. Bastian says:

    There are some services that can be used to “unshorten” links.

    I’ve given two different ones a try and http://www.unshorten.com/ just gives the full DOI-URI (So http://doi.org/dng becomes http://dx.doi.org/10.1093/BIOINFORMATICS/BTR050)

    http://unshort.me/ even resolves the DOI-URI and gives back http://bioinformatics.oxfordjournals.org/content/27/7/1013

    Both services have an API that’s easy to use.

  6. Sebastian Peters says:

    I don’t see any real benefit of shortDOIs. The length of a “normal” DOI is not a very strong argument (e.g. twitter uses t.co shortener). It just adds another redirect in front of dx.doi.org.

    Therefore there is no need for a special shortDOI->DOI service. Just speak HTTP :-)

    curl -D - -s http://doi.org/dvq | grep Location |cut -f 2- -d " "

    The biggest problem with shortDOIs: they are not DOIs – they are just a link to a DOI – and there is no statement about persistent of shortDOIs AFAIK.

  7. Martin Fenner says:

    Sebastian, thanks for clarification. And for the trick parsing the 301 redirect http response. I thought that shortDOI works differently, but it is really just a very special URL shortener.

  8. Pingback: ScienceCard named Finalist in Mendeley/PLoS API Binary Battle | Gobbledygook