Page:Finch Group report.pdf/42

 items downloaded in 2011 were published journal articles, but it is notable that they tended to be relatively old: the top two articles were published in 2001 and 2002 respectively.

4.20. For researchers in a number of disciplines, however, subject or discipline-based repositories are a more important part of the landscape: a place where they go for information, to see what’s new, to share early findings with their peers, and to look for collaborators, as well as to deposit their own articles. Provision is very patchy, and there are many gaps. But for researchers in a number of fields, subject-based repositories constitute an important element in their daily workflows.

4.21. Among the most notable of such services are ArXiv, predominantly but not solely for the physics community; CiteSeerx for the computer and information science communities; RePec for the economics community; the SSRN for the social science community more generally; and PubMedCentral for the biomedical and life sciences communities.


 * ArXiv is a preprint repository, for papers before they are submitted to a journal for peer review and publication. It contains c735,000 full text articles, and is growing at about 75,000 articles a year. There is minimal filtering of incoming papers for quality control purposes. There are about a million downloads a week.


 * CiteSeerx harvests documents and other material such as algorithms, data, metadata, services, techniques, and software; and it creates a citation index that can be used for literature search and evaluation. It has over 1.5 million documents with nearly 1.5 million unique authors and 30 million citations.


 * Research Papers in Economics (RePEc) is a collaborative service at the heart of which is a database of working papers, journal articles and software. In addition to working papers (which are disseminated among economists much more commonly than in most other research communities) it provides information about 692,000 journal articles, 629,000 of which are downloadable. But it does not itself host or provide access to the articles; rather, it provides metadata and links to documents it harvests from archives across the world. It estimates about 700,000 downloads a month.


 * The Social Science Research Network (SSRN) consists of a number of subject-based networks and encourages the early distribution of research results by soliciting and publishing submitted abstracts of research papers. It has agreements with a wide range of journals, publishers, and institutions. The SSRN eLibrary consists of an abstracts database of over 380,000 items and a collection of some 315,000 full text PDFs. It is widely used in the social science community, and has over 8m downloads a year.