On 27 and 28 September, the South African Department of Science and Technology (DST) convened a high-level two-day workshop on access to research data. The workshop was designed to address what South Africa’s response should be in relation to the OECD Declaration, Principles and Guidelines on Access to Research Data from Publicly Funded Research. A hint as to why this workshop was being convened now came from a press cutting included in the conference pack, reporting that South Africa is being considered as an additional member of the OECD, something that would be a major boost to the country if it were to come about. Another reason was mentioned by Owen Njamela, from the Chief Directorate, R&D Investments at the DST: that in the last few months the DST has announced a considerable increase in its strategic R&D targets for the next decade as a way of increasing the country’s international competitiveness. This means that the number of postgraduate degrees and the levels of research output will need to grow radically in the next decade. It was good to see these targets being linked to open approaches to knowledge and information sharing, in contrast to the restrictive and lock-down approach of the Draft Bill on IPR for Publicly Funded Research published for comment a few months ago (see my blog entry of 13 July 2007). What the workshop was after, Njamela said, was to establish what it would take to create a really effective data sharing system in South Africa.

Because I see this as an important event, I am going to blog in this post the key outcomes, decisions and forward planning that hat emerged from the workshop and then provide, here and on the OpeningScholarship project blog additional postings on the keynote speeches and the presentations from local speakers, as well as some South African case studies. Keynote speeches were by Paul Uhlir of the US National Academies of Science and CODATA, Bernard Minter, Chair of the World Data Centre System at ICSU and Professor of Geophysics at the Scripps Institute of Oceanography, University of California San Diego and Beatriz Torres, Programme Officer from the Global Biodiversity Information Facility.

The key understanding that emerged from the workshop was that, although there are a number of legitimate limitations on openness when it comes to research data – such as official secrets, personal privacy and the proprietary rights of private sector research – the default option, as spelled out by the OECD Guidelines, should be for open access and restrictions should be the exception and not the rule, only invoked with good reason. This is particularly important when data has been developed from publicly funded research. While locking up data in proprietary systems increases the fragmentation and cost and can become a barrier to the conduct of science, the keynote speakers argued that open access makes data available for use across disciplines and countries, allows for automated knowledge discovery, improves the potential for verification and accuracy and facilitates North-South and South-South transfer.

There are strategic reasons for ensuring that research data is properly disseminated and curated in South Africa. As the ASSAf Report on Scholarly Publishing in South Africa made clear, South Africa needs to increase its research visibility, needs to grow its output of high-quality publications and attract a younger cohort of scholars. And, as the ASSAf programme grows the output of local journals, I argued tin my presentation that there need to be links between scholarly publications and underlying data sets if the maximum benefits are to be gained from research investment. Looking forward, the trends are towards greater interactivity between scientific journal articles and the underlying data, for collaboratories and virtual workspaces, for the additional layers of interpretation that can be offered by semantically-rich XML documents, and for automated analysis, abstraction and correlation of data. Open Access makes this much easier.

An important issue for South Africa is the need to retain and grow the numbers of young researchers. The current system of evaluating scholars through their output of journal articles and the citation impact of these articles provides a disincentive for researchers who have grown up in a digital world, who expect rapid results in a collaborative global community of scholars and who recognise the need for high-speed supercomputing to access and analyse the vast and growing amount of data now available. Some participants felt that policy-makers are limping behind the development of new research approaches and that South Africa needed to become more forward-looking in its research policies – in fact the keynote speakers challenged South Africa to leap the technology gap to take its place at the forefront of developments.

Action plan

At the end of the workshop, Owen Njamela from the Chief Directorate, R&D Investments in the DST proposed an action plan,
based on the recommendations of a series of presentations and workshop sessions. The Department would:

• Explore the recommendation made by the workshop participants for an audit of skills, curricula, databases, and systems.

• Draft a narrative report from the meeting into a draft policy document/ guidelines for data management, access and reuse.

• Undertake internal departmental consultation to ensure awareness by all departments and the executive of the DST and to identify human capital.

• Consultation with other departments – Director-General forums and research forums within government and in the industry cluster system. (There is a framework of bilateral cooperation with other departments, Njamela said and it is also important to include government departments with their own science councils, like Mintek in the Department of Mining.)

• Consultation with universities/science councils (concurrent)

• Presentation of a policy to Cabinet (June 2008, when the government considers policy priorities).

• Funding considerations proposed to National Treasury

• Implementation (2008 financial year). There might be institutions active ahead of
  that implementation.