FAQ
Open Science III

When designing a dataset, we recommend following the established practices of your research discipline, particularly regarding its scope, structure, metadata, and data formats. It is also important to ensure that the dataset complies with the requirements of the Open Science III call, as described in the Rules for Applicants and Beneficiaries – Specific Part (Section 5.7.1).

If you are unsure which practices are considered standard in your field, we encourage you to join the EOSC CZ Working Groups and initiate a discussion with experts from your research community. The EOSC CZ Working Groups provide a platform for sharing experience, exchanging best practices, and building consensus on recommended approaches across different scientific disciplines.

Data stewards can come from a variety of professional backgrounds, including research data management specialists, librarians, IT professionals, and researchers. The key requirement is that they possess the necessary expertise in research data management and the FAIR principles.

Together with the Czech data steward community, we have identified three main data steward profiles, each with distinct roles and responsibilities. For project- or team-based data stewards, a good understanding of the relevant scientific discipline is an additional advantage.

There is no universal estimate for the workload required to FAIRify a dataset. The level of effort depends on the type, volume, quality, and current state of the research data, as well as on the extent of the FAIRification needed. Applicants are responsible for determining the proposed staffing and budget, which will subsequently be assessed by the evaluation committee based on the nature of the proposed dataset. We recommend consulting Open Science experts and data stewards at your institution or faculty when planning these activities.

FAIRification typically includes preparing research data for publication, cleaning and structuring the data, creating documentation and metadata, standardising data and file formats, addressing licensing and other legal aspects, performing quality control, depositing the dataset in a trustworthy repository with a persistent identifier (PID), and creating a metadata record in the National Metadata Directory (NMA). During the repository deposit process, researchers are usually supported by a data curator, particularly with PID assignment, licence selection, and metadata creation.

Yes. The determining factor is not how well the research data are already organised, but whether their preparation involves expert FAIRification activities. These may include, for example, designing and enriching metadata, standardising data and file formats, creating documentation, improving interoperability, defining licensing conditions, or depositing the data in a trustworthy repository. The outcome should be a FAIR dataset that provides added value for future reuse.

However, if the work consists solely of publishing data that are already FAIR-compliant or of their administrative registration without any further expert processing, the requirement for a non-trivial FAIRification process would generally not be considered fulfilled.

Simply redepositing an already published dataset in another repository is generally not considered sufficient to meet the requirements of the call. To provide added value, FAIRification should involve additional expert activities, such as enriching or refining metadata, creating or expanding documentation, converting files into standardised formats, translating documentation into English to improve interoperability, or extending the dataset with additional research-related files.

The scope of the required FAIRification activities and the associated budget are determined by the applicant based on the specific characteristics of the dataset. The appropriateness of the proposed staffing, budget, and FAIRification effort will subsequently be assessed by the evaluation committee, taking into account the nature and current state of the original data.

When selecting a repository, we recommend first checking whether a disciplinary repository exists for your type of research data. Ideally, it should be listed in re3data.org and meet the requirements for a trustworthy repository (for example, by holding the CoreTrustSeal certification or another recognised certification). The re3data registry allows you to search for repositories by research discipline, data type, and other criteria.

If no suitable disciplinary repository is available, a general-purpose repository may be used, such as Zenodo or the National Data Catch-all Repository. One advantage of the National Data Catch-all Repository is its integration with the National Repository Platform (NRP). Metadata describing deposited datasets are automatically harvested by the National Metadata Directory (NMA), eliminating the need to register them manually in the NMA.

If your dataset is not stored in a repository from which the National Metadata Directory (NMA) automatically harvests metadata, you can create a metadata record manually. Before doing so, you need to obtain access to the NMA. Instructions are available in the NMA documentation.

After signing in using your institutional identity, you can create a new record by entering the dataset's persistent identifier (DOI or Handle). If metadata are available through DataCite, they will be imported automatically. Otherwise, you will need to enter the required information manually, including the dataset title, authors and their affiliations, and the publication date.

A manually created record can be edited for a limited period of time. If the dataset is later deposited in a repository from which the NMA automatically harvests metadata, the manually created record will be automatically replaced by the metadata harvested from that repository.

There is no universal definition of a typical dataset in the humanities and the arts, as the nature of research data varies considerably across disciplines. Examples include digitised archival materials, image and audiovisual collections, text corpora, archaeological databases, language data, survey results, qualitative interview data, 3D models of cultural heritage, and other specialised research datasets.

The type, scope, format, and structure of research data depend on the specific characteristics of each discipline. For the Open Science III call, the specific type of data is less important than ensuring that the resulting dataset complies with the FAIR principles defined in the Rules for Applicants and Beneficiaries – Specific Part (Section 5.7.1).

There is no universal definition of a typical dataset in the life sciences and materials science, as the nature of research data varies considerably across disciplines. Examples include astronomical images, human or other organism sequencing data, environmental contamination data, digitised herbarium collections, and data describing the chemical or spatial structure of different materials.

The type, scope, format, and structure of research data depend on the specific characteristics of each discipline. For the Open Science III call, the specific type of data is less important than ensuring that the resulting dataset complies with the FAIR principles defined in the Rules for Applicants and Beneficiaries – Specific Part (Section 5.7.1).