Statistics Package for EPrints -- Harnad Stephen Clark 15 May 2006 13:38 UTC

-----Original Message-----
From: Stevan Harnad [mailto:harnad@ecs.soton.ac.uk]
Sent: Saturday, May 13, 2006 3:53 AM
To: eprints-tech@ecs.soton.ac.uk
Subject: Statistics Package for EPrints (fwd)

---------- Forwarded message ----------
Date: Sat, 13 May 2006 17:31:13 +1000
From: Arthur Sale <ahjs@ozemail.com.au>
To: AMERICAN-SCIENTIST-OPEN-ACCESS-FORUM@LISTSERVER.SIGMAXI.ORG
Subject: Statistics Package for EPrints

Apologies for cross-posting

Following Eloy Rodrigues' posting of an excellent statistics package for
DSpace (I strongly recommend it to DSpace users), it might be opportune to
remind users of GNU EPrints that a similar package has been available for
EPrints for about 18 months. It is downloadable from
http://eprints.comp.utas.edu.au:81/archive/00000262/.

The Statistics Add-on to the EPrints platform is responsible for the
gathering, processing and presentation of access, content and administrative
statistics generated by EPrints usage. It was developed at UTas = University
of Tasmania (Australia) based on principles established by The University of
Melbourne (Australia) and the package was subsequently enhanced by feedback
from the University of Otago (New Zealand). Authors/depositors respond very
favorably to access to statistics about their papers, especially if they are
required to deposit (a "mandate"). They then see a reason for depositing!

The system was designed to be adjustable to any environment, because the
components can easily be configured, changed or extended, to respond to
different information needs. The modules are written in php and work off the
archive logs with a few (minimal) hooks into the Eprints mySQL database,
mainly to retrieve document titles. The insertion of new links on the entry
pages and the metadata display pages are usual. All the code is open-source.

Important features of the UTas statistics system:

*         Almost real time processing of event logs (adjustable from once a
day - most of us - to more frequently)

*         Database based stored data (in a separate MySQL database)

*         Detection and processing of the country of origin of accesses,
with the local campus separated out from the local country due to its
different characteristics

*         Monthly download stats - allows detection of citation or
presentation events

*         List of documents by access frequency in 4 weeks, monthly, yearly
or total (customizable)

*         User-friendly graphics and tabular data, pretty realistic flags

For examples, see in alphabetical order Arizona
http://dlist.sir.arizona.edu/, Otago http://eprints.otago.ac.nz/, Rhodes
http:// <http://eprints.ru.ac.za/> eprints.ru.ac.za, or UTas
http://eprints.comp.utas.edu.au:81/. In all cases look for the Detailed
Statistics link on the entry page for which I give the URL, or browse for
any document and you should find a link on the metadata page. Hope you like
what you see and install it too. Let me know if you do.

Arthur Sale

Professor of Computing (Research)

University of Tasmania

Australia