« JISC CETIS gets a new look Web site | Main | Exploiting The Potential Of Blogs and Social Networks »

November 14, 2007

Repositories as Web sites (again)

There's a recurring (though somewhat occasional) theme on this blog about the need for us to reconceptualise the software components we currently refer to as repositories as being Web content management systems or just Web sites.  This is more than a change of label - it changes the way we think about them and the kinds of questions we ask about their deployment.

By way of example I note that there is currently a thread of discussion on one of the UK repository discussion lists stemming from the question "Do you know if it is possible to store metadata about an article, and hide that from public access, so that members of the public won't see it?".  Not that there's anything fundamentally wrong with the question you understand, but the way it is phrased and the tone of the ensuing discussion leave me somewhat cold these days.

In terms of Web content management I suspect the same kind of question would be along the lines of "How do I configure my system to expose different levels of information on the article Web page, depending on how public it is?".  This is a subtle difference in phraseology but the consequences in terms of how we approach the problem space are significant.  The focus would shift away from metadata (and, by implication, the OAI-PMH) towards Web site design, usability and information architecture.  And, in more general terms, this change of emphasis would re-focus repository discussions on things like accessibility, cool URIs, REST, Google sitemaps, search engine optimisation, microformats, tagging, RSS and/or Atom feeds, etc., etc.

Exactly the same topics the rest of the world talks about when they are dealing with making information available on the Web!

Now, you may argue that the way repositories work is totally wrapped up in the metadata they contain and what use is made of it - and I would completely agree with you.  That is true of any Web-based system, more or less.  But one doesn't think first and foremost about metadata when one is dealing with, say, Flickr.  One is much more interested in the functionality of the Flickr site and the way that images can be integrated with other Web services.  Yet in every respect Flickr is a repository service that manages my content and exposes it on the Web.  Yes, it is a service that is fundamentally based on metadata but the primary focus of our attention around it lies elsewhere.

Interestingly, I don't think I have ever seen a repository-oriented discussion about search engine optimisation - yet, surely, that is how most repository-held content is discovered?


TrackBack URL for this entry:

Listed below are links to weblogs that reference Repositories as Web sites (again):


A few months back I was at the Open Ed conference in Utah and a woman beside me uttered the acronym "SEO" and I responded back with an 'Ick!' because I associate the term with sleazy companies stuffing metatags full of crap to rise to the top of *any* search. To which she quite quietly murmered "Not all SEO is evil." This phrase bounced around in my head all week long, and I began to realize not only was she right, but that my not understanding the work I've done on repositories in relation to SEO is to truly reveal the un-webby chauvinism of those endeavors. Sorry if this is a non sequitar, it's been sitting in my head for weeks looking for the right outlet, and lucky you, your post triggered its outpouring. Cheers, Scott

@Scott Liselie I think you have nicely summarised the tensions between the acedemics, researchers and librarians who feel that ideas should be found based on their merits (and who regard search engine optimisation as a work of the devil) and people in marketing departments or those with a responsibility to maximise the impact of a department's expertise and intellectual property - who would be quite comforable with making use of ethical approaches to maximise exposure of ideas in a digital world.

I think this is an area that the digital library community does need to engage in - and would endorse Andy's suggestioin that we need a "repository-oriented discussion about search engine optimisation".

I quite like the idea of reconceptualising "repository" as "Web site" - it seems to be your reasons for doing this are to make repository folk think about and then apply usability and people-oriented design principles (and SEO) to their 'repository' offerings and increase repository usage, which sounds like a good thing.

But, and I'm coming from a technical background of course, there are two things here really:

* the public face of the repository (submission and dissemination - which could be "the Web site" but could also be a desktop application, or something in-between (Silverlight/Flex))

* the repository itself. (I don't like using "repository" generally as it means something else to data managers and is confusing). This is the thing that facilitates the public face (in the discussion that left you cold, Les Carr said that at the most abstract this would be just "database") and you can't live without it.

I'd be worried about dropping that from "repositories" because you stop people from thinking about the other side - the integration of the "repository system" with the institution's systems architecture and the development of the foundation from which all those good Web sites and services are built from.

I remember several "portal" projects which tried to think about usability, created beautiful(ish) interfaces and then went "oh, where to we get the data from?" and quickly the sparkly interface was dropped because it couldn't actually deliver anything useful. Not because it was a bad interface, but because, say, the finance system wasn't up to processing credit card payments.

Flickr wouldn't be flickr without some underlying structure and businesses like it are dynamic enough to develop their foundations as they need to. Institutions are not agile and so I think we need to be careful from distracting people from the boring foundations building ;-) by talking too much about the "Web site" until we have an institutional architecture that matches an Internet start up.

For what it is worth, I mentioned SEO and repositories in my talk at an RSP Professional Briefing day - the "Making Effective Use" talk at:
http://www.rsp.ac.uk/events/ProfBrief.php#brunel so we're starting to think about it! :-)

The comments to this entry are closed.



eFoundations is powered by TypePad