Metadata describes data by providing essential information that helps in its understanding. This can include administrative metadata such as the creator or creation date, etc, or more detailed information beyond the administrative ones e.g. how the data was produced like the used technique, the instrument’s manufacturer,... Metadata can be unstructured, which doesn’t adhere to a specific template, or structured, which follows a defined template. To enhance the findability and reusability of the data, it is important that metadata follows a specification defined by a metadata schema.

A metadata schema represents a template, which specifies the expected elements and how they are structured, such as attributes’ names, value types, rules, etc. Metadata schemas are actually files, which can be represented in various formats like JSON, XML,... Using metadata schemas enables users to describe data with rich and structured metadata, enhancing the findability and reusability of the data.

A metadata document is a file including a digital description of data. It provides information such as the resource type, author, creation date, used technique, instrument’s manufacturer, etc. Metadata documents can be unstructured, which does not follow a specific template, or structured, which adheres to a defined template. Creating structured metadata documents, following a metadata schema, enhances the findability of the data and supports the reproducibility of results.

While a metadata schema represents a file defining the structure and rules for metadata such as attributes definitions, a metadata document is a file containing the values for such attributes, describing specific data.
N.B.: A metadata document should be valid against the defined metadata schema, this means it should follow the rules described in the defined schema.

The FAIR principles are a set of guidelines aimed at making data Findable, Accessible, Interoperable, and Reusable. To make data FAIR, it is essential to describe it with rich metadata, assign unique and persistent identifiers to (meta)data, register it in searchable repositories, and ensure it can be retrievable through standardized communication protocols. FAIRness is not a binary state (yes/no), but it reflects rather how effectively (meta)data fulfill the FAIR principles. This can vary depending on different criterias and the specific needs. An example of detailed guidance on applying the FAIR principles can be found here.

Unstructured metadata documents can be created by writing a free-text file in a 'readme'-style format using a text editor. Alternatively, structured metadata documents can be created by filling in the values of attributes following a defined template. It is recommended to provide structured metadata documents following metadata schemas, as this improves the findability and reusability of data. Various online tools are available to assist in the validation of such structured metadata documents.

There are published metadata schemas which are well-established and widely accepted by the user community, called metadata standards. Adopting existing standards avoids proliferation of descriptions. Different interdisciplinary metadata standards exist, e.g. Dublin Core and DataCite. There are also discipline-specific standards like NeXus, which describes neutron, x-ray, and muon experiment data and is also used in condensed matter physics and materials science. You can search for metadata standards in dedicated metadata registries, e.g. DCC Metadata Standards, FAIRsharing.org, RDA Metadata Standards Catalog, and others. Two instances of metadata registries including metadata schemas are provided: one hosted by the NFDI-MatWerk consortium, and a second one hosted by the NEP and JL-MDMC consortia.

You can start by creating a "readme"-style metadata document. Once you have a draft, feel free to contact us, we will support you in formalizing it into a proper schema. It can be considered as a good practice through broad adoption. However, to be recognized as a standard, it must be endorsed by a standardization body.

This information represents your metadata and can be obtained from various sources: it may be filled by memory, retrieved from physical lab notes, or extracted from existing resources such as data files, Electronic Lab Notebooks (ELNs), etc…

You can either create the metadata document using a text editor by reviewing the attributes defined in the metadata schema and entering the corresponding values or use a form generated from the schema to enter the metadata. Using a form makes the process easier and reduces the risk of errors. An online JSON editor is available to generate a form based on a JSON schema. eXact lab offers an open-source desktop application called Metadata Editor that provides a user interface, allowing users to easily create metadata documents with the help of forms, generated from already stored metadata schemas. These schemas are available in two different instances: one provided by the NFDI-MatWerk community and the second one offered by NEP and JL-MDMC communities.

Yes, you can map automatically the metadata included in the data files to metadata schemas. If further metadata is required, this can be added manually. In order to achieve this, you can either create your own customized mapping extension or use an existing tool that supports your data formats and used instruments. We offer a mapping service that extracts metadata from various data-files, and maps it to project-specific JSON metadata schemas ensuring better interoperability. Two instances of the tool are available: one hosted by the NFDI-MatWerk community and a second one by the NEP and JL-MDMC communities. A detailed documentation of the service including the supported instruments, input file formats and vendors is available here.

Data and metadata can be stored together in the same location or separately in different locations. This depends on the storage capabilities and your specific requirements. To align with the FAIR principles, your metadata should be available even if your data is no more accessible and it should be properly associated with the data it describes. Therefore, choosing the appropriate storage location is very important.

Data can be stored in different locations like Electronic Lab Notebooks (ELNs), databases, filesystems or data repositories. Storing your data in repositories ensures that administrative metadata are created and stored. In addition, your data becomes better searchable and findable. This contributes to more efficient data management and alignment with FAIR principles. There exist general-purpose data repositories (such as Zenodo), which support a wide range of disciplines. Additionally, there are also registries which include different research data repositories (such as re3data). Institutional repositories (such as KITopen) are available and managed by universities to store and publish their outputs. Project-specific data repositories are also available, such as the NFDI-MatWerk Data Repository hosted by the NFDI-MatWerk consortium.

If you only need to store administrative metadata, this can be done when uploading your data to a repository, as all repositories support this functionality. However, if you want to include more detailed metadata to enrich your data documentation, there are dedicated metadata repositories available for this purpose. Project-specific metadata repositories are available, such as the NFDI-MatWerk Metadata Repository hosted by the NFDI-MatWerk consortium and the MetaRepo Metadata Repository hosted by the NEP and JL-MDMC consortia.

No, you can choose whether you want to make your (meta)data public, keep it private, share it only with your colleagues or publish it with view-only rights. Most repositories provide the access control permission and user authentication functionalities in order to manage the access rights. In order to fully benefit from the FAIR principles, it is recommended to make your metadata publicly available even if your data is not.

Yes, you can still register your metadata in a metadata repository even if the data itself isn’t stored in a repository. It’s essential to link the metadata to the data it describes. This helps ensure the metadata remains accessible, even if the data becomes unavailable, which supports alignment with the FAIR principles.

Regardless of where your data and metadata are stored, establishing a clear link between them is essential. This association connects the descriptive information (metadata) with the actual object being described (data). This can be done by specifying the location of the data within the metadata document. The best way to do it is to use a globally unique and persistent identifier such as Digital Object Identifier (DOI), Open Researcher and Contributor ID (ORCID),...

You can search for data through general-purpose repositories (such as Zenodo), registry of data repositories (like re3data), or institutional ones (f.e. KITopen) depending on where it has been stored. Most of these repositories provide search interfaces allowing you to filter results using administrative metadata attributes. The access to the data may require authentication, depending on the permissions set by the data owner. We provide two search interfaces, allowing users to find specific created metadata that goes beyond the administrative metadata. The first one is hosted by the NFDI-MatWerk consortium and a second one offered by the NFFA and JL-MDMC consortia.

To enable the legal reuse with a proper citation of your data, it is important to publish it under a clear license (e.g. Creative Commons License, Apache License,...). A license is a legal agreement between the data creator and the end-user, or the repository where the data is stored, specifying how the data may be used. Choosing an open license or one with minimal restrictions ensures that others can reuse and cite your data in a legally compliant way.

No, without a specified license, you don’t have the legal rights to reuse the (meta)data. In this case, you can contact the author or data owner and ask them to provide a license.

Yes, most repositories provide the access control permission and user authentication functionalities in order to keep your (meta)data private or share it only with some collaborators during active research. There is also the possibility to publish your (meta)data with view-only rights, enabling controlled access. In order to fully benefit from the FAIR principles, it is recommended to make your (meta)data publicly available.