"Be conservative in what you do; be liberal in what you accept from
others".
Absolutely - completely agree.
UKOLN is taking forward the development of the harvesting and
aggregation service which underpins the Intute Repository Search. We
are developing what I call a 'heterogeneous metadata store', which
while using OAI-PMH and OAI-DC as it's default harvesting mechanism/
standard, will not demand further quality assurance or compliance form
source-repositories. Our philosophy is that we'll take whatever we can
get, and then we'll process what we have for particular services,
according to their needs. If a service has a requirement for a
stricter compliance with some norm or standard, then we can filter for
them accordingly. This is, in fact, the approach we already take for
the Search service, while delivering a differently filtered set of
data to other projects/services.
We will be investigating the harvesting of other metadata formats,
possibly using other mechanisms. Currently our baseline requirements
are:
- we only harvest and aggregate metadata records in well-formed XML
- we only harvest from repositories listed in OpenDOAR (we use
OpenDOAR as a service component)
- we only harvest 'open access' records.
err - that's it!
This is likely to remain the case for the foreseeable future, although
we're working this out now.
Paul
On 23 Mar 2009, at 12:20, Chris Rusbridge wrote:
> It is a bit strange, isn't it! My first thought was Intute
> repository Search (such an obvious and euphonious name, don't you
> think? See http://www.intute.ac.uk/irs/). They apparently cross-
> search 95 repositories, which doesn't sound a lot. There doesn't
> seem to be a way to suggest a repository to them, other than getting
> it into OpenDOAR.
>
> They also say "You need to make sure that the repository is
> completely compliant with the OAI-PMH protocol. Our harvesting
> mechanisms are quite strict, and although your repository may
> harvest with less strict harvesters, it may fail if it does not meet
> all the conditions for the IRS harvesting". Hmmm. That rather seems
> to break one of the Internet foundations: Postel's Law "Be
> conservative in what you do; be liberal in what you accept from
> others" (see http://blog.dshr.org/2009/01/postels-law.html). People
> do take OAI-PMH so seriously, when it's only a tool! Are there not
> important repositories that don't talk PMH, but that researchers
> might like to search?
>
> OpenDOAR does bang on about research, but at least gives a page to
> register (http://www.opendoar.org/suggest.php), and also offers a
> reasonably broad approach to contents, including datasets. Your's
> might turn out to be one of the few (apparently) UK repositories
> that really does do data!
>
> Of course, you could say, what the hell, who searches from these
> origins anyway. Link from your home page and a few other places and
> let Google etc do the rest.
>
> My suggestion? Do both...
>
> --
> Chris Rusbridge
> Director, Digital Curation Centre
> Email: [log in to unmask] Phone 0131 6513823
> University of Edinburgh
> Appleton Tower, Crichton St, Edinburgh EH8 9LE
>
> The University of Edinburgh is a charitable body, registered in
> Scotland, with registration number SC005336.
>
>
>
> On 23 Mar 2009, at 10:37, Michael Emly wrote:
>
>> In a number of postings to this list earlier this month,
>> acknowledgement was made that "research" repositories now need to
>> take account of material in formats other than text. The argument
>> typically is that research outputs can include data files,
>> modelling and various other related objects.
>>
>> While I have no quarrel with such statements, to my mind they
>> represent a rather narrow view of the management of information and
>> information resources within the HE environment. It is perhaps
>> easiest to illustrate this by citing specific examples from LUDOS,
>> the multimedia repository being developed for the University of
>> Leeds (https://ludos.leeds.ac.uk/). The first thing to say is that
>> you will not find any research papers in LUDOS - those are
>> currently held in the consortial-run repository White Rose Research
>> Online http://eprints.whiterose.ac.uk/ (based on Eprints, which is
>> a platform specificallly designed for this type of material).
>>
>> However what you will find in LUDOS is a number of collections
>> which specifically support the research process. 2 prime examples
>> are:
>> 1. The Timescapes "collection". This is an archive of primary
>> research material (records of interviews, etc.) associated with a
>> major longitudinal study into "the dynamics of personal
>> relationships and identifies".
>> 2. Virtual pathology (not yet live). A collection of virtual
>> pathology slides from the early 20th century to provide a "rich
>> educational and research resource of diseases".
>> In both cases, the explicit intention is that the materials
>> generated through the research process should support further
>> research activity and also be available to support student learning
>> (both directed learning and research projects) where appropriate.
>>
>> Alongside such "research"-oriented collections, there are others
>> which have been:
>> 1. created explicitly for student learning such as digitised music
>> scores
>> 2. created to support the University's internal processes such as a
>> collection of documents to support EKT activity
>> 3. created in order to more generally foster both research and
>> learning/teaching (L&T) such as digitised page images of the
>> medieval illuminated manuscripts held in the University Library.
>>
>> The final report of the MIDESS project(funded by JISC and RLUK)
>> commented in 2007 on the dangers of establishing too rigid a
>> distinction between materials for research and for L&T within the
>> repository context and argued strongly for a more integrated
>> approach (http://ludos.leeds.ac.uk/midess/MIDESS-final-report.pdf p.
>> 20-21). Sadly, there has been too little evidence of that to date,
>> at least from what I have seen.
>>
>> These thoughts arose from a desire to register LUDOS with an
>> appropriate OAI-PMH harvesting service in order to make our
>> holdings discoverable within a national/international framework,
>> but I am having difficulty in identifying one which is not focussed
>> solely on research outputs. Can anybody help?
>>
>> Many thanks
>>
>> Michael
>>
>>
>> ********************************************
>> Michael Emly
>> Collection Management Services Team Leader
>> Leeds University Library
>> tel. +44 (0)113 343 6444
>> email: [log in to unmask]
>> Postal address:
>> Brotherton Library, University of Leeds, Leeds LS2 9JT
>>
--------------------------------------------
Paul Walk
Technical Manager
UKOLN (University of Bath)
http://www.ukoln.ac.uk/
[log in to unmask]
+44(0)1225383933
--------------------------------------------
|