We have only one journal here at the Electronic Text Centre (University of New Brunswick) who have taken up Crossref. While I'd like to say that they did so because of things like bibliographic impact and an eye towards the future of citation cross referencing (and they very well may have considered these things to be fringe benefits), I think they mainly caved to peer pressure. Since the publication is a science journal, and many other journals in their field were using DOIs, they felt as though they should also have them.
The journal pays $275 a year for their membership to Crossref.
The process for encorporating crossref into journal production certainly isn't trivial, but I also don't think it's as hard as everyone seems to think it might be. First of all, members of Crossref are, more or less, obligated to check the bibliographies of their upcoming articles against Crossref's database in order to flush out any DOIs those articles might have. There's a few ways journals can go about this. Primarily, people seem to recommend a program called "eXtyles" which will take your bibliography, transform it to XML, check it against any number of indexes, and extract DOIs for linking. I'm sure it does other things. It also costs over $20,000 or something ridiculous. I believe it's typically used by large publishing groups like the NRC Research Press.
However, Crossref itself has a small web-based version of eXtyles on their site that you can operate for free. In my experience it works fairly well with well-formatted biliographies. http://www.crossref.org/SimpleTextQuery/
So, our journal will run their references through that web-form and publish the resulting DOIs as part of the references. That's step one.
Step two is your own DOI. This involves assigning a DOI prefix (if you're using OJS 2.2.3 or above, you can have it generate this DOI for you) to the articles. After your articles are published - in order for CrossRef to begin indexing your articles - you have to send Crossref an XML file (one, or many, depending on how you publish). OJS can export Crossref XML from the Import/Export menu under Journal Manager.
I'd like to say that you can just log in to Crossref and submit your XML at this point, but the export won't necessarily catch everything you need to look out for. I'd advise opening that XML up and checking for validity. Crossref uses their own fairly intuitive DTD for their XML. You might want to double check your metadata as well.
Once this is all taken care of, then you can upload to Crossref. Their user interface is incredibly non-user friendly and it's hard to tell if you're actually doing the thing you're supposed to be doing. Their emails back to you don't help either. This is the email I received that indicated an error in the XML I submitted.
- Code: Select all
<doi_batch_diagnostic status="completed" sp="svc2.crossref.org">
<msg>DOI: atlgeol.2009.004 , does not contain '/'</msg>
I know that this is basically telling me that I didn't have a prefix for my DOI, but these sorts of things might not be clear to everyone. In any event, once you get the file up in crossref then you can actually appreciate what the service offers.
For what it's worth (and despite the effort), I think that CrossRef is a great service. If you take the time to make those DOIs clickable, you can have entire bibliographies as clickable links (assuming those articles have been indexed). Furthermore, your journal would get more traffic from those referencing its material.
If anyone has any other questions about DOIs I can attempt to answer them. I basically became responsible for this stuff by default when the journal decided to pay the bill and asked us to sort it out. I'm certainly no expert.
Users browsing this forum: Google [Bot], Yahoo [Bot] and 1 guest