On Open Access Analytics in Europe

        Karsten Schuldt (FH Graubünden)      analytics · open access

One of the work packages in the SONAR project is concerned with “Open Access (OA) Analytics”. This is a field on which work has been done already in the last years in other European countries. SONAR can profit from this work.

There is no unifying definition, aim or structure for OA analytics, but there are trends:

  1. Definitions on OA are still open for discussion. Usually the classification of unpaywall is used, because that is a main source of data for most projects in OA analytics. But there are still several proposals for other classifications.
  2. The aim of OA analytics is not always the same. Research funders, for example, often focus on questions of “compliance” to OA-policies, but other – and more interesting – questions keep getting asked (e.g. who publishes in OA, the workflow of libraries / OA-offices).

Looking at the solutions for OA analytics that have been proposed in the last years, one finds that there was a time when mostly “one-shot studies” were conducted, but now more and more nations have moved towards forms of continuous analytics. Most, but not all of the “one-shot studies” use data from Web of Science or Scopus that they buy, normalise and analyse to give pictures of compliance towards OA policies at one point in time. Today, different countries try to find other (and better) data sources. They have built – or are in the process of building – infrastructures for continuous analytics using other data sources.

Forms of continuous OA analytics

One can identify four kinds of approaches:

  1. Based on “current research information systems” (CRIS) (Czech Republic, Slovakia, maybe Austria in the future), where researchers of all academic institutions of one nation are expected to report all of their publications.
  2. Based on input or reports from academic institutions (Denmark, The Netherlands), where universities are expected to provide e.g. lists of researchers. In Denmark, the OA-indicator is based on an annual process which is shown here: https://www.oaindikator.dk/oa-docs/annual_production_cycle_en.pdf
  3. Based on automatic harvesting of data (France), where efforts are made to do OA-analysis on the basis of open data.
  4. Based on open and non-open data sources (Germany), which – until now – seem to give the best insights into the development of OA: https://open-access-monitor.de/

SONAR context

All of these approaches include a lot of manual work to normalise and analyse the data used. No solution has yet provided a way to reduce such work completely. As such, every solution needs continuous funding. In some countries (like Denmark), the OA-analysis is tied to a state agency which provides this funding, others are still funded as projects and have yet to find a sustainable solution. In the case of the SONAR project we try to learn from those examples and put forward a working strategy towards continuous OA analytics in Switzerland.