I had the pleasure of meeting Rob Sullivan in person at our first Working Open Workshop in Berlin. Rob has been leading development on PMC-ref, one of our community run open source projects for science. PMC-ref is an API that checks an article’s references to see which ones are free to access in PubMed Central, an archive of biomedical and life science articles.
Rob loves learning, bibliometrics, and building APIs which makes this passion project will suited for him! I interviewed Rob to learn more about PMC-ref, open access, and how you can help during our Global Sprint 2016, June 2-3.
What is PMC-ref and what inspired you to start it?
PMC-ref is an API that analyses an article’s references and determines which ones are free to access in PubMed Central.
My scientist friends had told me about the amount of time it took “following the paper trail” when doing their research so it seemed like something could be automated. On the way to working on that I also thought it would be cool to know which references are free.
Why is open access important?
Much like how open source software opened up software development to a greater number of people, me included, I think open access can do the same for research. And for selfish reasons, I’d like more people finding things out so I can read about it. ;)
Has working outside of academia helped you appreciate the importance of open access?
Yes, as a student I could access everything I needed via the University library but that’s not the case now. My job doesn’t depend on accessing research though but more generally speaking I understand the frustration at obstacles to carrying out work.
How will being able to check if an article’s references are free to access help researchers?
In several ways I hope. One idea is that the researcher can quickly evaluate how much research they can access for free should there be a need to pay for an article. Kind of like buy-one-get-x-number-of-articles-that-I-might-need-to-read-free. Another is that it’s possible to build up a whole network of research that can be accessed and then be able to share that network (a feature perhaps for someone to implement in the Global Sprint?)
It’s also possible to use the API to search by journal so a researcher could browse for their favourite journal and check what’s been archived in PubMed Central.
What problems have you run into while working on PMC-ref?
Working out how to collect and present the data took some time and I soon realised some of the technical decisions I made early on as I was learning about software development turned out not to be the right ones so there’s a limited number of articles that can be checked. The goal now is to make it a dynamic look up to PubMed, which will give the benefit of looking up citations (as long as they’re indexed by PubMed).
And then there was working out how to build an API that made sense, this took some trial and error and I definitely violated the DRY principle to get it working.
The front-end, although basic, was a challenge. It has to work on old browsers and across multiple devices and with low bandwidth so the page couldn’t be bloated. Still working on this one.
What kind of skills do I need to help you build PMC-ref?
PMC-ref is built using Node.js so knowledge of or desire to learn JavaScript.
CSS is a highly sought after skill.
Ideas! I’d love to hear how the API could be used and/or improved to make it more useful.
Oh and if someone knows about how to migrate data from a document store to relational database that would be fantastic.
How can others help you during the Mozilla Science Global Sprint, June 2-3?
There’s this great starter issue to implement a new endpoint to return citation data and this would really make PMC-ref amazing (but would I need to change the name? : / ).
Bonus question: Where is your Research Fox sticker?
The Research Fox has jazzed up the baby changing table!
—
Come join us wherever you are June 2-3 at the Mozilla Science Global Sprint to work on PMC-ref and pick up your own Research Fox! Have your own project or want to host a site? Submissions are open for projects and site hosts.