Monitoring compliance with Open Access mandates | Gold Open Access Infrastructure

"Research funders and universities have adopted a range of policies in support of OA. Some are relatively permissive in how grant-holders or employees can comply with the policy, whereas others are more precise in what is required. As yet, there is no widely-used vocabulary for common elements in an OA policy, such as allowable embargos, required licences, or specific repositories into which material should be placed. This not only makes an author’s job more difficult, something that the Sherpa-FACT service tries to address for UK funder policies, but also makes funders’ and universities’ jobs more difficult as they investigate how effective their policies have been. It is probably impractical for each research paper to be flagged with its compliance (or not) with the full range of OA policies that might be relevant to it. This is especially true as some publishers, universities and funders might make bespoke bilateral arrangements. However, some parts of the puzzle are being established that will help. We might think about these in terms of how they help funders and universities in some increasingly typical scenarios: [1] A funder wishes to know the extent to which research papers from a particular grant have been made OA, according to a certain definition of OA. In this case, the FundRef service from CrossRef will help, allowing published research papers to be associated with particular funders. The outcomes from the NISO steering group on OA metadata, and the related V4OA vocabularies project, should enable published research papers to be identified as OA. Of course, Creative Commons licences already do this, when they are used. Thus, in theory, the effectiveness of a funder’s policy might be judged as the ratio of OA papers (either tagged as OA, and/or using CC licences) to all papers tagged with that funder’s name at FundRef. [2] A university wishes to know the extent to which researchers based there are complying with both relevant funder policies, and with its own OA policy. Clearly, some of the infrastructure from (1) above is also relevant here. Also, though, there need to be reliable ways to identify particular universities. There are a number of lists of research performing organisations in the UK, and no doubt similar lists exist elsewhere. For example, the UK funding councils keep a list of institutions that qualify for them as 'higher education institutions', and this list is also used by the Research Councils, who also keep lists of eligible research institutes, the UK access management federation has a list of member organisations, various Government departments and agencies have lists, eg the UK Home Office list of educational institutions trusted to sponsor migrants to the UK (PDF). The CASRAI initiative offers a possible way to come to a consensus on how to identify universities for the purposes of this scenario, and a UK group is being set up for this purpose. Many OA policies require research outputs to be made available under certain terms and conditions, for example with respect to machine readability or re-use rights. These conditions can be expressed in various ways, some for example by the licence used. The Public Library of Science has developed a useful spectrum of 'openness'd, and helped develop a pilot tool that interrogates websites to estimate how open a research paper is. Unsurprisingly, given the attention on OA in the UK since the Finch Report, there is a lot of activity related to OA policies, and a lot of discussion over what they should be and how they should be implemented. This is driving some perhaps overdue innovation in the information ecology that supports research, information management and publishing, which is all to the good. Nevertheless, some lightweight coordination of this innovation might benefit everyone."