It is widely recognized that sharing data is beneficial not only for science but also for the common good, and researchers are increasingly expected to share their data. However, many researchers are still not making their data available, one of the reasons being that this activity is not adequately recognized in the current reward system of science. Since the attribution of data sets to individual researchers is necessary if we are to include them in research evaluation processes, the purpose of this paper is to explore the feasibility of linking data set records from DataCite to the authors of articles indexed in the Web of Science.
DataCite and WoS records are linked together based on the similarity between the names of the data sets’ creators and the articles’ authors, as well as the similarity between the noun phrases in the titles of the data sets and the titles and abstract of the articles.
The authors report that a large number of DataCite records can be attributed to specific authors in WoS, and the authors demonstrate that the prevalence of data sharing varies greatly depending on the research discipline.
It is yet unclear how data sharing can provide adequate recognition for individual researchers. Bibliometric indicators are commonly used for research evaluation, but to date no large-scale assessment of individual researchers’ data sharing activities has been carried out.