Draft eResearch diagram/map

Today we’re having an eResearch strategic [UPDATE: strictly speaking IT & Research] planning day at the University of Western Sydney. The meeting has been organized by IT Services, which is a really healthy sign of a commitment to taking eResearch seriously.

But what is eResearch?

The usual definitions say something about “Advanced IT and communications technology in used for research” and some add stuff about computation, collaboration, cross-discipline, and data-driven research. I thought, I’d try to map this out for today’s meeting in this diagram.

Reading from the bottom to the top the idea is that there are some basic eResearch enablers that need to be in place. The Australian Government has recognized this and is funding national infrastructure, but each institution obviously needs to have some of its own.

In the middle applications layer there are some services that we all need, that are more-or less similar across disciplines, we need to be able talk to each other, have mechanisms for sharing data, and the organization needs to know about who’s doing what.

It’s well understood that a lot of eResearch services are not generic , so that’s the top cloud, where effective eResearch help needs to a be a partnership between researchers and tech-specialists (sometimes they’re the same people).

I’ll update this with feedback from UWS, interested in what others think. Use the comments.

(ORS means Office of Research Services)

Copyright Peter Sefton, 2011-12-19. Licensed under Creative Commons Attribution-Share Alike 2.5 Australia. <http://creativecommons.org/licenses/by-sa/2.5/au/>

  • Chris Rusbridge

    Peter, this looks pretty good to me. I think what’s missing (it may be in “embedded services” but I don’t think so) is the equivalent of “bio-informatics” for the various disciplines, ie a discipline-oriented computation and data-transformation function at the top. This will often pull in a wide range of techniques, and is in a sense anti-generic, ie somwhat at war with the generic IT layer lower down. Rsearchers often want, nay need, capabilities that are beyond the bleeding edge.

    • http://ptsefton.com ptsefton

      Yep – I agree. I think maybe this is sort of implied by the ‘virtual lab’ label on the top level but it needs to be explicit – I need to work out how to show how disciplines cut slices through this model.

      Sam Searle @datalibsam noted on Twitter that discipline specific repositories might be needed at the top layer as well.