Petermr’s blog recently had a post about the how institutional repositories don’t deal well with dynamic scientific data. The post can be found at: http://blogs.ch.cam.ac.uk/pmr/2011/08/21/talks-data-journals-and-datasharers-for-science/. This shouldn’t be taken the wrong way. In a previous post, Petermr’s Blog outlines several criteria for a good repository (http://blogs.ch.cam.ac.uk/pmr/2011/08/19/criteria-for-successful-repositories/). However, this blog post reminds us that institutional repositories are great for static digital objects with which an institution would like to preserve or mange over time for the benefit of that institution. When it comes to scientific data, however, much of it is not static. Researchers have needs for versioning, distributed repositories, ability for both long and short term storage, terms of use, more ways to share the data and citations, etc. With this in mind, institutional repositories are unable to handle this type of dynamic approach to data. More often than not, most institutional repositories are unable to handle the size of many of the datasets out there created by scientists. This is not to say that institutional repositories should be replaced with something different and better. Definitely not! Institutional repositories have their place in many institutions and supply a much needed service for static objects that need to be preserved and managed for an institution. However, with the advance in dynamic data and datasets, there needs be another type of service for researchers. Some institutions have already gone in this direction. The Dataverse Network is one such project from the partners of DATA-Pass (http://www.icpsr.umich.edu/icpsrweb/DATAPASS/about.jsp). It will be interesting to see what other “Data-Share” projects pop up to respond to this need.
Celeripedean 