“Publishing 2.0” was an unconference session at this year’s dotastronomy conference, and FigShare was one of the new tools discussed in this session. In a nutshell, FigShare is a free online repository for scientific results from all disciplines. Created in 2011 by then-PhD student Mark Hahnel on Mediawiki software, it is now supported by Digitial Science. Apart from figures – the name is a confusing misnomer – FigShare allows the sharing of posters, papers, videos, audio files, datasets, and more. There are no rules or constraints, everything can be uploaded. Sharing and getting credit for all research outcomes – not just the ones which eventually end up in peer-reviewed papers – is the main selling point for FigShare.
The main advantage of FigShare is speed. A figure can be uploaded and made publicly available within seconds. Compared to regular scientific publishing this feels like switching from plate tectonics to a Ferrari. At the same time, FigShare attempts to maintain some of the functionality of an online journal. It offers long-term archiving, metrics (number of views and shares), and provides persistent handles as well as DOI identifiers, which make all FigShare posts citable. If you always wanted to cite a video of the UV Sun in a paper about coronal pulses, FigShare is perfect. These benefits make FigShare a very attractive tool for open science workflows.
Over the past few months, FigShare has slowly gained traction. One of the interesting aspects of this process is to watch how FigShare can be used for entirely new models of scientific publishing. The most prominent example is probably the collaboration with ‘Faculty of 1000′, a review platform for biology and medicine, which now also offers its own open online journal ‘F1000 Research’. FigShare hosts complementary material to papers published in this journal, which are directly embedded in the paper text with user-friendly widgets.
Several groups routinely use FigShare as a complementary publishing tool, somewhere between the regular papers and an open labbook and usually combined with dissemination via social media (for example, Ethan Perlstein at Princeton, Lorena Barba at Boston University). In particular, FigShare is becoming a hub for the sharing of the types of research output that are usually not published at all – posters, PhD theses, ‘dark data’ (invisible data, for example from failed experiments), micro-results. Even grant proposals have been published and disseminated on FigShare. My own uploads to FigShare are plots that may or may not appear in a future paper, results from student projects, null results that still may be useful for other people, or figures that could be interesting to a wider audience. Another option is to use FigShare for public outreach, for example by uploading a figure that explains a specific finding in an upcoming paper. The possibilities are practically endless; in a world which has only one accepted currency (the paper), FigShare looks like Europe before the Euro, only less inconvenient.
So far, astronomy is extremely underrepresented on FigShare – the category ‘astrophysics’ received only 10 submissions in August and less than 100 in total. The top post in astronomy is the output from the Galaxy Zoo project, which received almost 1200 views and 30 shares. To give an idea of more typical numbers, the figures I have uploaded to FigShare are usually viewed about 200 times, mostly within a week or two. Interestingly, this is comparable to the average number of downloads for my peer-reviewed papers on ADS – over much longer timescales, of course.
FigShare is still in its early-adopter phase. Currently it feels like arriving too early at a party – you can have a chat with the host and get to try out everything, but it is still pretty lonely and your friends haven’t arrived yet. You are wondering if they will ever show up. With only 11 submissions I have already the most ‘views’ on FigShare among astronomers and I am easily in the top five among all users. At this point, the best use of FigShare may be to simply experiment with new ways of publishing and sharing of scientific outcomes. Based on the feedback it is easy to guess that most of the ‘views’ on FigShare come from people interested in astronomy, but only a minor fraction from professional astronomers. Maybe this just means that the best application of FigShare in its current phase is indeed public outreach.
To attract more users and to play a larger role, FigShare needs to solve some its current problems. Although posts are organised in categories and tagged by the users, the functionality to search the FigShare archive is still skeletal and cannot compete with ADS and arXiv, although the just released API promises to be an important improvement. This would become more of an issue if the number of submissions increases. Another problem is that the sources of views and shares are not transparent (i.e. we don’t see the referrals), and citations are not tracked yet on the site. It will be interesting to follow the further development of the site as the next improvements are implemented.