The number of repositories is growing fast worldwide but still many of them do not have their own domain or subdomain, and for this reason it is not possible to add them into our analysis. Some institutions maintain several databases with completely different URLs which penalize the global visibility they have.
We are still unable to add usage/download statics but there are many initiatives already working on standardization of the collecting methods, so we expect that global data should be available soon.
Following several requests we now show two global Rankings. One that covers all repositories as was shown in previous editions (Top 300), and a new one that focus only on Institutional Repositories (Top 300 Institutional).
There is a minor change regarding the calculation of the number of rich files as in this new edition we are again using formats other than pdf (doc, ppt, ps) to obtain the data. Contrary to the methodology we used to make the other Rankings, the figures for rich files are combined and not treated individually.
The French HAL central repository, and its subsets like INRIA, Social Sciences and Humanities (HAL-SHS) or IN2P3, are at the top of the institutional repository list.
Important repositories like PubMedCentral, CiteSeerX and Smithsonian/NASA Astrophysics Data System, do not use standard suffixes to design their contents (e.g. papers in acrobat format with file names which extension is not .pdf). This is a bad practice as it reduces the visibility of these documents to the search engines.
Our policy is not to include collectors or metarepositories, with one exception, which is DiVa, the interface that the Uppsala University provides to more than 20 Nordic repositories. Many of these institutions do not have their own systems but link their contents to the DiVa portal. Unfortunately, this means that many of the papers are under different domains and hence they do not contribute to the DiVa’s rank.