We use several sources to collect and evaluate academic scientific publication on a country-wide scale, and we apply it to the case of France for the years 2015–2020, while presenting a more detailed analysis focused on the reference year 2019. These sources are diverse: databases available by subscription (Scopus, Web of Science) or open to the scientific community (Microsoft Academic Graph), the national open archive HAL, and databases serving thematic communities (ADS and PubMed). We show the contribution of the different sources to the final corpus. These results are then compared to those obtained with another approach, that of the French Open Science Barometer for monitoring open access at the national level. We show that both approaches provide a convergent estimate of the open access rate. We also present and discuss the definitions of the concepts used, and list the main difficulties encountered in processing the data. The results of this study contribute to a better understanding of the respective contributions of the main databases and their complementarity in the broad framework of a countrywide corpus. They also shed light on the calculation of open access rates and thus contribute to a better understanding of current developments in the field of open science.

Research infrastructures (RIs) offer researchers a multitude of research opportunities and services and play a key role in the performance, innovative strength, and international competitiveness of science. As an important part of the generation and use of new knowledge and technologies, they are essential for research policies. Because of their strategic importance and their need for significant funding, there is a growing demand for the assessment of their scientific output and impact. Current research information systems (CRIS) have contributed for many years now to the evaluation of universities and research organizations. Based on studies on the application of CRIS to infrastructures and on a recent French report on the scientometric assessment of RI, this paper analyzes the potential of CRIS and their data models and standards (in particular the international CERIF format and the German RDC model) for the monitoring and evaluation of RIs. The interaction between functional specificities of RI and standards for their assessment is outlined, with reference to their own potential to stimulate and share innovation in the networks located inside and outside RIs. This societal challenge, more than an academic issue, is on the way to further harmonization and consolidation of shared and common RI metrics.