Hi Claudia,
very interesting topic. In a desperate attempt to keep my answer as concise
as possible, I'd like to sketch two common cases implemented by our clients
and some general observations afterwards. I'll see if there's anything I can
add to your wiki page.
*Case 1: Calculating publication KPI's within DSpace and serving them to
SAP, that works as the institution's ERP system.
*
For a large comprehensive university (+20k students, +5k researchers) we
assisted in implementing enhancements that would disambiguate journal names
on publications in DSpace, together with a staff DB integration that
disambiguates author names. Also journal impact factors and citation
information got integrated on the DSpace items, mainly from ISI.
DSpace acts as the controlling source over publication data and as far as
the institution is concerned, if your (metadata) record is not in DSpace, it
doesn't exist for purposes like promotion applications and annual reporting.
Instead of just allowing export of "raw" publication lists, several custom
indicators are offered. One example being, if I remember correctly, average
impact factor for the top-20 publications (by impact factor) for the last 2
years. This approach required that there's a clear distinction between
affiliated publications and publications not affiliated with the
institution. Although the repository policies allow that a researcher builds
his complete personal bibliography in there, it's required for the reporting
procedures of the institution that only affiliated publications are
considered for the indicators.
*
Case 2: Integration with a CRIS system in which the repository only serves
as a web front-end for public data*
Almost all of our clients that work with different CRIS systems (Pure,
symplectic, converis, ...) have in common that the repository is no longer
the "controlling source" for the publication metadata. This often means that
the DSpace submission process is no longer used for entry of the (meta)data
but that this is being taken care of through the web user interface of the
CRIS system. Also very often, the CRIS systems have no functionality for
offering online access to the full text, and even if the have it, there's
often a lack of fine grained access controls or embargo's (like DSpace has).
So what you see is that the metadata gets pumped automatically from the CRIS
into DSpace, and that full text is being attached within DSpace.
In some cases, the DSpace submission process is still in use for types of
content that aren't considered in the CRIS.
*
A few observations*
- While DSpace is currently able to deal with publication data in most
cases, it's very challenging if you would want to implement grant & project
information. I know almost of no installations that attempted to build
full-fledged CRIS functionality in DSpace.
- Sometimes metdata is being controlled/edited in different sources. As a
result, there sometimes arises a need for complex update and/or
synchronization scripts. My very personal opinion is that you can avoid a
lot of complexity by deciding which source/system will be the "controlling
source" for the (meta)data, after which it will be exported or visualized in
other systems. Pushing through an update of an item in one direction between
systems is manageable, building a bi-directional update process is tedious
and can go wrong for a variety of reasons.
- Very often, the barriers to use DSpace for institutional reporting
purposes is non-technical but organizational. Politically, things can get
very tricky if a library would invoice/account a cost to the research
management office for some of the repository costs. Sadly this can lead to
silo's with similar data in different departments.
with kindest regards,
Bram Luyten
@mire
Esperantolaan 4 - 3001 Heverlee - Belgium
2888 Loker Avenue East, Suite 305 - Carlsbad, CA 92010 - USA
atmire.com - Institutional Repository Solutions
2011/6/30 Claudia Jürgen <[email protected]>
> Hello all,
>
> at the moment we are working on an institutional bibliography.
>
> There are a variety of systems/scenarios possible. The frontiers between
> repository, cris, publication management, collaboration tools are
> transient.
>
> As I've found no detailed comparison of systems apart from a spreadsheet
> by Dorothea Salo, I started one and put it along with some of my notes
> on the DSpace Wiki
> https://wiki.duraspace.org/display/~cjuergen/Institutional+Bibliography<https://wiki.duraspace.org/display/%7Ecjuergen/Institutional+Bibliography>
>
> The system comparison is selective and proprietary. There are a lot more
> interesting systems/solutions around.
>
> Feel free to enhance the comparison and add your notes.
>
> Furthermore I would be interested in those already running an IB on the
> reasons why they chose the system they are using now.
>
> Thanks in advance for your input
>
> Claudia Jürgen
> --
> Claudia Juergen
> Universitaetsbibliothek Dortmund
> Eldorado
> 0231/755-4043
> https://eldorado.tu-dortmund.de/
>
>
> ------------------------------------------------------------------------------
> All of the data generated in your IT infrastructure is seriously valuable.
> Why? It contains a definitive record of application performance, security
> threats, fraudulent activity, and more. Splunk takes this data and makes
> sense of it. IT sense. And common sense.
> http://p.sf.net/sfu/splunk-d2d-c2
> _______________________________________________
> DSpace-tech mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/dspace-tech
>
------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2d-c2
_______________________________________________
DSpace-tech mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/dspace-tech
