Setting up an institutional archive: some experiences

From: Eriksson J=F6rgen <>
Date: Tue, 10 Feb 2004 16:04:34 +0000 (GMT)

Dear all,

There has been some discussion earlier about the estimated cost for
institutional archives. We will give a summary of how we have approched
the task of setting up an institutional archive, some of our experiences
and costs related to that work.We also hope this might stimulate others
working with institutional archives to share their experiences with us.

We have a long way to go to get self-archiving as a natural
part of a researchers publishing behaviour. At the moment we
have about one hundred plus full text documents in our repository
[]. It also serves as a bibliographic
database for research papers, reports, articles etc. written by
researchers affiliated to Lund University, and so far most of our records
are just bibliographic. The way we have approched this task gives us a
rather slow start in terms of resources in the archive in the short term,
but we hope that it will be rewarding in the long run.

We started working by setting up a demonstrator that showed a unified
university perspective. We made a survey of what was available already
at departmental web sites etc. and added samples from those to have some
content to show as examples. We "launched" the demonstrator at a half-day
seminar on electronic publishing/scientific communication in May 2002.

We held five seminars on self-archiving and oa-issues in general
during 2002 and 2003, all which have been fairly well attended (20-40
researchers). We plan to hold at least one seminar per semester on
different aspects of open access/institutional repositories. This
semester we have also managed to be part of a range of courses
which the Learning and Teaching Development Centre at Lund University
[] offers researchers. We also
held a seminar in late 2003 for department and faculty librarians with
the aim to enable them to promote LU:research and to work as first-line
support to their researchers when questions on using LU:research arise.

The ignorance among the researchers on the broader developments in
scientific communication is widespread, but there is an awakening
interest that we find important to support by arranging these seminars
and courses. An introduction to copyright issues which are part of
our seminars is very appreciated and we have extended that work to
include the creation of a standard licensce agreement that includes the
right to self-archiving and re-use of papers in electronic versions of
dissertations. This work is done together with the faculty of Law and
the university's legal department. We have a finalised document now
[] and we have also been
discussing the creation of a web-site run by the legal department where
researchers can find copyright-related information and personal support.

Conclusions from our work.
Different perspectives.

* the university perspective only does not get the departments/researchers
interested enough to participate. We have identified 3 different
perspectives that the repository will try to satisfy:

1. The University:

The single, unified, entry point is the University perspective
[ ]. It was also the University that already
from the beginning wanted the adding of bibliographic records to the
repository, not just full text. In the longer term the University sees
LU:research as a marketing tool, and when the researchers/departments
really start to use it, as a tool to help assessing research activities
at different departments. There are no central, formal decisions on the
use of the repository at the moment in this decentralised university. Our
library director is raising awareness on different university management
levels about these issues and a proposal for a policy will be on the
agenda this spring.

2.The faculty/department:

One of the results of the first seminar was that we were approached
by the information committee of the medical faculty. They wanted to
show the output from their research but where not interested in just
showing it in the university context. Together with them and their
faculty librarian we created an independent user interface to their
sub-set of our institutional repository, Lund Virtual Medical Journal
[]. There are mainly bibliographic records
so far, but now that we have gained their confidence and given them
something that they feel belong to just them we have started to talk about
self-archiving and through the support of their library they have begun
adding full text to the bibliographic records. Here we would like to point
out what a great help the RoMEO site is to that work. We have also talked
to a LU researcher who is the editor of a medical journal and he has
just recently obtained permission from the publisher (Taylor & Francis)
to self-archive all articles by authors from Lund in LU:research. T&F
does not accept self-archiving in general according to the RoMEO list.

3. The Individual researcher:

We are working with a software engineering research centre in our second
"pilot project". They will use the same method as the medical faculty
to generate a view of their own publications on departmental level. The
addition here is that we will create a simplistic way for the individual
researcher to dynamically add his publication list to his/her personal
home page.

= this generates extra technical work in creating ways of delivering
the records in ways that suits our users and also in setting up their
user interfaces when they don't have enough technical support of their
own. You could argue that these costs are unnecessary, but as pointed
out above, to be visible only in a central university service does not
seem to be attractive enough to give an incentive to participate.

Need for creating awareness.

* researchers in general are indifferent/unaware of "the crisis/new
possibilities in scientific communication" but there is an awakening
interest and fanning that interest through seminars, lectures
etc. generates support for our local repository.
= this takes time and will generate a cost although it is not connected
to the technical costs

* ignorance regarding copyright is wide-spread and information on these
issues are much appreciated (we are also lucky to have a member of the
faculty of Law who is interested in these matters AND who is also a very
good lecturer :-)) = this takes time and will generate a cost although
it is not connected to the technical costs

Our marketing strategy after finishing the pilot projects is in short:

We are approaching every department, mainly through their libraries,
and offer to visit and show through our examples/pilots what we can do
for them. We already have a couple of departments beside the pilots who
have started adding records/full text to the archive.
Strengths we will stress:

* Impact. Their output will reach more collegues through OAI-sevice
providers. We have created a set for "full text available in archive"
to make it possible for OAI-harvesters to harvest only those records. The
problem here is the lack of good (not experimental) OAI service providers
besides OAIster.

* The economies of scale in centralised administration and data
management/security that still leaves the possibillity to have their
own user interfaces at personal and faculty/departmental levels.

Practical stuff that is also needed.

Helpdesk functions. Although we find the administration module in the
GNUeprints software satisfactory , our experience is that not every user
find it totally intuitive.
Metadata quality. Time is also spent on controlling the metadata in the
records added to the archive, something we believe is needed, if the
OAI service providers are going to work with some amount of precision.
This kind of work will increase with the number of users adding
records. It is done either by us centrally or at department/faculty level.

As of today Lund University Libraries, Head Office are spending ca 1 FTE
on the central administration/maintenance/marketing of LU:research. To
this should be added the quality control/support performed at
faculty/department level, which will depend on the level of ambition at
each unit.

We mean that it isn't a correct analysis to say that the cost of archiving
is trivial, at least not when looking at the actual situation at Lund
University. Hopefully it will be a more self-generating system in the
long run, but seminars and courses about publishing issues will probably
be needed for quite some time to come.

We are very interested in getting in touch with other institutional
archives to share experiences on ways of implementing an institutional
archive. Please send comments and your experiences either to the list
or directly to us.

Best regards,
Jorgen Eriksson and Sara Kjellberg
Lund University Libraries Centre for Electronic Publication
Received on Tue Feb 10 2004 - 16:04:34 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:20 GMT