Abstract

Data interoperability, or interoperability in general, has become a key requirement for the research community, agencies, and institutions worldwide. We think and move more towards FAIR data principles, with the idea that people should be able to have access, explore, and understand the structure and content of research activities, outputs, and related data.

To this purpose the research community is trying to set data standards, but given the diversity of people, careers and environments is this really possible? Does moving towards a set of standards really help push development in research or does it limit the scope of what we seek to achieve?

We strive for data interoperability, but while it is admirable to try and contain all the information we need in one standard it is probably impossible to achieve this. Different people following different career paths in different research environments have different needs, and these needs evolve and change in time. Based on these needs CRISs are trying to fulfill a variety of use cases that keep growing in number and specificity, making it more and more difficult to set one (or even a handful) of standards that are effective enough to cover everyone’s needs.

Today the research community (in line with government and funding agencies and repositories) is still trying to adopt standard protocols such as OAI-PMH and common standards for data exchange such as CERIF, Dublin Core, or OpenAire. In line with this, CRISs are focusing on supporting these standards and, more in general, moving towards a unified model that can facilitate the transfer and usability of data. However, not all needs can be satisfied by a single set of standards. Data formats such as CERIF or Dublin Core, although guaranteeing a standard for baseline interoperability, do not seem to cover the requirements of (all) repositories. This results in the creation of different metadata schemas that can 'fill the gaps', such as RIOXX or the XML-based MODS and DIDL schemas. In addition, different countries have different requirements that are determined by governments, funding agencies, or other national agencies, which in turn leads to the development of country-specific formats.

So is setting standards a double-edged sword, and how should we go about it? We must understand what objectives we are trying to achieve through the standards, such that they cater to the different needs of institutions, government agencies, or other stakeholders.