The main aim of the Ranking Web of Repositories is not really to rank them, but to promote Open Access initiatives and global access to academic knowledge. Following that objective we promote good practices, especially those intended to increase the visibility and usage of the OA contents.
Our point of view is that an Open Access Institutional Repository, including the full texts of the documents, is the TREASURE of the university, not another bibliographic databasel. We support the development of publicly open CRIS, an important tool for the information management of the university outputs. But mixing the bureacratic aims with the mission of the university repository, ie to increase the access to and the visibility of the knowledge generated by the university, is not the correct strategy.
The Ranking is also a powerful tool for penalizing bad practices, with especial emphasis in the awful naming proposals by software developers that ignore librarian traditions and in many ways are going against intellectual rights of depositing authors.
We truly believe the lengthy and useless addresses coined for the repository items have a negative impact in their web visibility and affect the authors will to deposit as it makes difficult the citation of full texts in future papers.
In previous editions we penalized the incorrect naming of the filetypes, such as not using pdf suffix for the acrobat files. Unfortunately, the way in which Google Scholar recognizes this situation was far from perfect, so many repository managers does not recognize the true purpose of filtering by filetype but a methodology error.
Proposal as updated 11th September 2014
As the pdf issue is only a minor part of a problem, we decide to fix it and to focus on the more important topics. We are opening a public discussion on these issues and the following proposals are being updated according to the comments we recieve before next edition of the ranking scheduled for late January 2015:
- Calculation of Scholar indicator will no longer be filtered by pdf (Jan15 edn)
- Institutional repositories that don’t use the institutional domain will be excluded (Jan15edn). There will exceptions for small organizations in developing countries with limited resources that need to be hosted externally
- Pseudo-institutional repositories of academic subunits will be excluded. (Jan15edn). This refers mostly to Individual School or Faculty repositories if parent institution has already a central one
- Repositories using ports others than 80 or 8080 (Jan15edn). A short list of additional ports will be considered on request
- Institutional repositories that use the name of the software in the host name will be excluded. (Only new entries). it intends to give stability to the URLs if the software is changed in the future. We strongly recommend the use of meaningful names like 'repository', 'digitalprints' o similar ones. We discourage the use of confusing/misleading ones like 'dlib' or 'archive'
- Institutional repositories that use more than 4 directory levels for the URL address of the full texts will be excluded. (extra time is needed, delayed to Jan16edn). Software designers need time for finding a feasible solution.
- Institutional repositories that use more than 3 different numeric (or useless) codes in their URLs will be excluded.(extra time is needed, delayed to Jan16edn). Software designers need time for finding a feasible solution. The following is a living example of informative, citable, good SEO option http://www.amazon.com/Institutional-Repositories-Environment-Information...
- Repositories being part of a CRIS will be excluded (Jan15edn). CRIS are not requiring full text deposit and not designed for properly bibliographic management of the records. Motivation for supporting green road to Open Access can be seriously affected.
- Repositories with more than 10% of the records not linking to OA full text versions willl be excluded (Jan15edn). Items with restricted access due to embargoes are OK.
- Repositories without author profiles combining article-level metrics will be excluded (proposed Jul15edn). A best practice model (not the url) is in http://citec.repec.org/p/b/pbl23.html
- Institutional repositories including items not authored by its members exceeding the 10% of the total are now already excluded, but this policy will be applied in a more strict way. Third-party contributions to hosted conferences, textbooks or special projects are allowed.
For example, it will be acceptable:
On the contrary, these are not:
Other examples of bad practices:
http://repository.domain.edu:8080/dspace not redirecting to http://repository.domain.edu/
Similarly with .../jspui/ or .../xmlui/
Comments to the editor [email protected]