I didn't suggest that tweaking the UIs of IRs is enough (though it might
be a good idea too). In fact, what I was suggesting is that we can keep
the IRs more or less as they are AND provide a central service using
their content (through harvesting) in line with what you are suggesting.
This is an architectural change in presentation but not in deposit. The
Web 2 functionality could be (principally, at least as regards this
discussion) in that new central service rather than in IRs. That doesn't
stop us carrying on with our method to provide the content that will be
neccessary for any service to work effectively. We might need to make it
clear to users of each IR that the content was being used in a new
attractive central service, where they could bookmark it, use feeds and
so on, as appropriate. That way they would know they needed to deposit
in local IRs for it to be able to provide the services it offered.
For those of us who are slightly less knowledgeable about the
technicalities of OAI-PMH, could you please say in more detail why it is
"service-oriented" rather than "resource-oriented" and why this matters?
Thanks.
If institutional repositories have a "poor fit" with the social
networking of academics, presumably any Web 2 features could remedy that
by providing the functionality that would more in keeping with it.
However, it does seem more sensible for social networking to be
non-institutional, which could be fitted into a centrally harvesting
service as above.
I welcome your last remark about repositories having a role to play. If
we accept that they don't do they social networking part but they are
nonetheless the only useful content providers, could we proceed by
talking about how centrally harvested services could be better
integrated into the overall architecture to provide those new functions?
I think we can accept that the distributed model works for getting
content but not for social networking.
Does this work? It would seem like an eminently sensible idea to use
different tools for different jobs, but to capitalise on them by using
them in a co-ordinated manner. That way we would address not only
deposit, but also uptake once the deposit is there, providing an
incentive to engage with repositories and any new service using them.
Thanks,
Talat
-----Original Message-----
From: Repositories discussion list
[mailto:[log in to unmask]] On Behalf Of Andy Powell
Sent: 11 March 2008 11:10
To: [log in to unmask]
Subject: Re: Central versus institutional self-archiving
Just to clarify again, I am *not* suggesting that it is sufficient to
tweak the user interfaces of current IRs in some way, nor that we need
to make them more Web 2.0-like.
I can't stress this enough.
My argument is fundamentally an architectural one. The architectural
approach that we have adopted - a natwork of institutional repositories
with an aggregation layer on top, using the OAI-PMH as glue - does not
appear to be working.
Further, I suggest that two of the reasons why the current architectural
approach is failing are that:
1) our choice of glue was wrong (though we didn't know that at the time)
because it turns out that the service-oriented approach of the OAI-PMH
does not fit well with the resource-oriented approach of the Web
architecture
and
2) that institutional repositories have a poor fit with the social
networks of researchers and that it is therefore difficult to realise
any social network effects.
These are architectural issues first and foremost.
Now, there are, of course, other factors, many of which are beyond my
detailed understanding - not the least the reluctance of researchers to
do anything! And I am happy to concede that mandates almost certainly
have a role to play in this space in the future. But unless we start
asking questions about our architectural approach we are always going to
be fighting a battle against the infrastructure that we have created.
I think that we have to acknowledge that repositories play a significant
role in the social networks of researchers and that to understand the
architecture we have to understand its relationship to the way that the
scholarly social network functions (or not!).
Andy
--
Head of Development, Eduserv Foundation
http://www.eduserv.org.uk/foundation/
http://efoundations.typepad.com/
[log in to unmask]
+44 (0)1225 474319
> -----Original Message-----
> From: Repositories discussion list
> [mailto:[log in to unmask]] On Behalf Of Talat
> Chaudhri [tac]
> Sent: 11 March 2008 09:55
> To: [log in to unmask]
> Subject: Re: Central versus institutional self-archiving
>
> Despite my earlier comments today, to clarify, there seems no
> reason why we cannot have both a new UI that makes the
> content more attractive AND follow the mandated IR model.
> This is only the same as saying, for instance, that we can
> pursue advocacy and voluntary engagement AND at the same time
> seek institutional mandates (or integrate mandatory staff
> reporting with a software management system and front-end
> repository, which comes to the same thing). It seems to me
> that you are saying precisely this, if I understand you correctly.
>
>
> Talat
>
> -----Original Message-----
> From: Repositories discussion list
> [mailto:[log in to unmask]] On Behalf Of Steve Hitchcock
> Sent: 10 March 2008 20:08
> To: [log in to unmask]
> Subject: Re: Central versus institutional self-archiving
>
> A note of clarification. The terms Web 2.0 and social
> networks span a multitude of services, so just considering
> those services referred to already in this discussion:
>
> Google, Internet Archive
> No new content, harvested services
>
> Slideshare, Flickr, YouTube, blogs
> New digital content for which there was previously no
> coordinated means of dissemination (taking the generic cases,
> i.e. photos, slides, rather than YouTube vs Google Video, etc.)
>
> IRs, subject repositories
> New content, purpose OA; established alternative means of
> dissemination for the primary target content, but those means
> not open access (mostly)
>
> The differences may be a little clearer now. To make these
> comparisons meaningful we have to look at their role in new
> content generation and author motivations. It's not enough to
> decide models are successful for one type of content,
> especially when we haven't measured the relative successes,
> and to try and replicate that for other content on the basis
> of labels. The analysis has to be deeper than that.
>
> There are lessons to learn for repositories, particularly in
> terms of improved services, perhaps interfaces too, but first
> we have to understand the type of target content and look at
> the available evidence. Is there is a UI solution to IRs that
> can beat mandated IRs for OA content (the few we have of
> these so far)? The clues above and the evidence seem to be saying not.
>
> Steve
>
> At 13:16 10/03/2008, you wrote:
> >Hmmm... the fact that you "have never, ever, ever heard
> anyone refuse
> >to use our institution's timetabling software because the user
> interface
> >isn't good enough" rather misses the point - or my point at least.
> >
> >This is not a discussion about whether the user-interface of
> each IR is
> >good enough or not.
> >
> >It's a discussion about what makes one or more repositories
> grow into a
> >viable scholarly social network. The UI is a small facet of that...
> >what I'm suggesting is that the 'social networking' aspect is more
> >important and that we need to understand that aspect rather
> better than
> >we do now in order to understand why repositories remain unfilled.
> >
> >Take something like Slideshare (www.slideshare.net) as a
> case study -
> >albeit one with significant differences to the scholarly
> repositories
> >space in terms of content, responsibilities and the surrounding
> >political landscape of scholarly publishing. But bear with me
> >nonetheless...
> >
> >Ask yourself what makes Slideshare such a successful repository of
> >presentation-like material - i.e. such a compelling place to surface
> >that sort of content on the Web? Yes, part of the answer lies in UI
> >type issues. But more fundamentally the answer lies in the network
> >effects of a globally concentrated service. Could the functional
> >equivalent of Slideshare have emerged by getting people to put their
> >presentations on the Web in a distributed manner and then harvesting
> >them into a central service? I don't think so. Ditto Flickr, ditto
> >YouTube, ditto ...
> >
> >Having said that, I accept that the blogsphere is a good
> counter case
> >study... because the blogsphere does give us an example of a healthy
> >social network built on a distributed based of content,
> using globally
> >concentrated services (Technorati, et al.) that harvest that content
> >into multiple single places. The interesting question is what makes
> >these approaches work (or not) and what we can learn from
> them to help
> >fill our repositories (centralised or distributed) without relying
> >solely an a "thou must deposit" type approach.
> >
> >But as I said on eFoundations... imagine a world in which every
> >institution mandated to their academics that they must only
> blog using
> >an institutional blogging service - would that support or hinder the
> >development of a vibrant academic blogging environment?
> >
> >And before you ask, I wouldn't mandate that people deposit in a
> globally
> >concentrated service either - for me, the only mandate that
> matters for
> >OA is one that says that scholarly output must be surfaced openly on
> the
> >Web.
> >
> >Andy
> >--
> >Head of Development, Eduserv Foundation
> >http://www.eduserv.org.uk/foundation/
> >http://efoundations.typepad.com/
> >[log in to unmask]
> >+44 (0)1225 474319
> >
> > > -----Original Message-----
> > > From: Repositories discussion list
> > > [mailto:[log in to unmask]] On Behalf Of Leslie Carr
> > > Sent: 10 March 2008 10:30
> > > To: [log in to unmask]
> > > Subject: Re: Central versus institutional self-archiving
> > >
> > > On 10 Mar 2008, at 09:55, Stevan Harnad wrote:
> > >
> > > > Brewster Kahle may have the disk space, but if his is to become
> the
> > > > global database, then why should individuals have local websites
> at
> > > > all? They could all set up shop in the Global Wayback
> > > Machine -- or,
> > > > for that matter, store directly in Google, saving it the trouble
> of
> > > > having to harvest!
> > >
> > > Brewster or Google can do all they like - if the content
> ain't there
> > > it can't be harvested. People often think that somehow
> > > "repositories"
> > > are failing, but they're no different from "web sites" in that
> > > respect. An examination of research and university web sites show
> > > that researchers have out-of-date, incomplete pages and
> sometimes no
> > > pages at all. My own Head of School's home page is just
> in the form
> > > of an FTP listing of some files he occasionally puts
> there. Others
> > > of my senior colleagues have home pages that are over three years
> > > old and miss out on describing an entire generation of
> projects and
> > > their outputs.
> > >
> > > The fundamental problem is not repository software, it is
> > > researcher's disinclination to disseminate. And I am
> convinced that
> > > the repository software isn't fundamentally at fault
> because I have
> > > never, ever, ever heard anyone refuse to use our institution's
> > > timetabling software because the user interface isn't good enough
> > > (though it is appalling), or because it doesn't integrate
> into their
> > > personal calendar (which it
> > > doesn't) - they just get on and use it because it does a job they
> > > need to do.
> > >
> > > But that isn't to say that we at won't be working our hearts out
> > > trying to make EPrints better and more functional!
> > > --
> > > Les Carr
> > >
>
|