<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
</head>
<body bgcolor="#ffffff" text="#000000">
<font face="Helvetica, Arial, sans-serif"><small>Hello,<br>
<br>
I have to prepare a study about the usage of the documents available
into <a href="http://www.ifremer.fr/docelec/">Archimer</a>, our
Institutional repository. <br>
<br>
Here are some of the first results: <br>
<br>
</small></font>
<ul>
<li><font face="Helvetica, Arial, sans-serif"><small> There is now
more than 3700 documents available freely (about 2000 publications, 500
reports, 60 books, 900 conferences proceedings). 46% of these documents
have been published after 2000. These documents are mainly about life
sciences but there are also some documents about physics, Geology… </small></font></li>
</ul>
<font face="Helvetica, Arial, sans-serif"><small></small></font>
<ul>
<li><font face="Helvetica, Arial, sans-serif"><small>It seems that
90% of the full-texts are downloaded directly from Google. Then, the
success of a document is of course linked to its quality but also to
technical aspects: </small></font></li>
</ul>
<ul>
<li><font face="Helvetica, Arial, sans-serif"><small>Documents that
have a large number of pages (ex: theses) are more downloaded that
document that have a few pages (ex: publications)</small></font></li>
</ul>
<ul>
<li><font face="Helvetica, Arial, sans-serif"><small>It seems than
Google does not index documents that are bigger than 10Mo. Then
documents bigger than 10Mo are then downloaded 10 less times than other
documents</small></font></li>
</ul>
<ul>
<li><font face="Helvetica, Arial, sans-serif"><small>Even if the
documents are downloaded from Google, we can detect a high proportion
of universities and scientific institutions in users.</small></font></li>
</ul>
<ul>
<li><font face="Helvetica, Arial, sans-serif"><small>Documents that
are only available into our repository are, of course, more downloaded
that documents that are available in many places. This is particularly
true for publications. Recent international publications, that are also
available from editor web site, are less downloaded that old
publications that are only available from our repository.</small></font></li>
</ul>
<ul>
<li><font face="Helvetica, Arial, sans-serif"><small>It seems that
the first one that put a document have a kind of bonus from Google. For
example, when we put a publication on line before the editor, it seems
that it is more downloaded, even after the publication on the Editor
Web site. When we load a publication several months after it was
available on editor web site, it can hardly be found on Google. </small></font></li>
</ul>
<ul>
<li><font face="Helvetica, Arial, sans-serif"><small>In march 2008
(march is usually the higher month for stats), the documents were
downloaded in the following way: </small></font></li>
<ul>
<li><font face="Helvetica, Arial, sans-serif"><small>Reports
average: 20 downloads (min : 0, max: 200)</small></font></li>
<li><font face="Helvetica, Arial, sans-serif"><small>Theses
average: 50 downloads (min : 0, max: 300) </small></font></li>
<li><font face="Helvetica, Arial, sans-serif"><small>Publications
average: 8 downloads (min : 0, max: 70)</small></font></li>
<li><font face="Helvetica, Arial, sans-serif"><small>Conference
proceedings average: 15 downloads (min : 0, max: 100)</small></font></li>
</ul>
</ul>
<font face="Helvetica, Arial, sans-serif"><small><br>
</small></font>
<ul>
<li><font face="Helvetica, Arial, sans-serif"><small>The average rate
of downloads have decreases since last year. Indeed, the documents
loaded after July 2007 match about 44% of the total documents available
into our repository. However these 44% documents match only 22% of the
downloads during September 2008. this is a little bit disturbing and I
am not sure of the causes:</small></font></li>
<ul>
<li><font face="Helvetica, Arial, sans-serif"><small>There are more
and more documents available on the internet, so maybe it is more and
more difficult for a document to be visible</small></font></li>
<li><font face="Helvetica, Arial, sans-serif"><small>Maybe there is
a kind of a self saturation. For example, 1000 of the 37000 documents
are about oysters (which is one of the main research subjects in our
institute). There are clearly some documents (ex: annual survey
reports) that are quite similar, so there are not as interesting as
there was all original search results.</small></font></li>
<li><font face="Helvetica, Arial, sans-serif"><small>Even if the
full text are indexed quite quickly by Google, maybe it take some time
before the new documents appear in top of google list.</small></font></li>
</ul>
</ul>
<font face="Helvetica, Arial, sans-serif"><small><br>
Could you please tell me if you have the same comments on your own
repository?<br>
</small></font>
<ul>
<li><font face="Helvetica, Arial, sans-serif"><small>What is the
average download rate per type of documents in your repository?</small></font></li>
<li><font face="Helvetica, Arial, sans-serif"><small>Does it increase
or decrease when the number of documents is increasing?</small></font></li>
<li><font face="Helvetica, Arial, sans-serif"><small>…</small></font></li>
</ul>
<font face="Helvetica, Arial, sans-serif"><small><br>
Thanks for your help.<br>
<br>
Kind regards,<br>
Fred</small><br>
<br>
<br>
</font>
<p class="MsoNormal"><span style="" lang="EN-GB"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="" lang="EN-GB"><o:p> </o:p></span></p>
<div class="moz-signature">-- <br>
<font face="Arial" size="2">Fred Merceur<br>
Ifremer / Bibliothèque La Pérouse<br>
<a class="moz-txt-link-abbreviated" href="mailto:frederic.merceur@ifremer.fr">frederic.merceur@ifremer.fr</a><br>
Tél : 02-98-49-88-69<br>
Fax : 02-98-49-88-84<br>
<a href="http://www.ifremer.fr/blp/">Bibliothèque La Pérouse</a><br>
<a href="http://www.ifremer.fr/docelec/">Archimer, Ifremer's
Institutional Repository</a><br>
<a href="http://www.ifremer.fr/avano/">Avano, a marine and aquatic OAI
harvester</a><br>
<br>
<b>Avant d'imprimer, pensez à l'environnement!</b>
</font></div>
</body>
</html>