Improving ecological (meta)data FAIRness through semantic services: integration of EcoPortal in LifeWatch Italy new platforms

The Data Portal is a data repository, based on DSpace, that provides FAIR data and metadata. It helps scientists to share their (meta)data and also to re-use data created by others. The data schema is based on the Darwin Core standard and controlled vocabularies. The metadata schema associated with each dataset is the Ecological Metadata Language profile LifeWatch (EML 2.2.0; Vaira et al., 2022).

The LifeWatch Italy Metadata Catalogue is an information management system based on GeoNetwork 4.2.2, designed and implemented to enable access to several resources from a variety of providers through descriptive metadata, enhancing and promoting the information exchange and sharing among organizations and researchers. The LifeWatch Italy Metadata Catalogue gives access to different resources and their metadata that are based on two main standards:

• ISO 19139 (VREs, services, workflows and research sites);

• EML 2.2.0 (datasets).

Each metadata profile is organized in sections, which reflect the main information related to each specific resource. Each section contains optional and mandatory metadata elements (Vaira et al., 2022).

One of the objectives of the last implementation was to provide a new version of the Data Portal and the Metadata Catalogue and also integrate both with EcoPortal. The integration aimed to use well defined and agreed concepts/classes for the metadata attributes and its values.

EcoPortal uses an advanced version –collaboratively developed and experimented in FAIR-IMPACT– of the OntoPortal technology. It supports the scientific community in the creation and management of semantic artefacts and their use to harmonize (meta)data. The recent updates of EcoPortal have enabled to improve the metadata schema (MOD 2.0) associated with semantic artefacts and to align it with other catalogues and repositories (Tarallo et al., 2024). In addition, the integration of the Ontology FAIRness Evaluator (O’FAIRe) tool, allows to assess the level of FAIRness of semantic artefacts through a metadata-based automatic FAIRness assessment methodology (Amdouni et al., 2022).

Challenges and implemented solutions

The challenges we aim to address involve harmonising metadata and data in our platforms using semantic artefacts, while ensuring that semantic artefacts in use are FAIR so that others, humans or machines, can find, access, interoperate and reuse them.

The integration encompasses the implementation of a REST API connector, through HTTP GET calls, that links the editing tools of Data Portal and Metadata Catalogue with EcoPortal, thus enhancing the management of data and metadata to ensure FAIR compliance.

As for DataPortal, two REST API access points (classes and properties) are used to extract the hierarchy of selected semantic artefacts. Those artefacts are subsequently converted and saved within the platform in DSpace proprietary format, ensuring a continuously performing system. The update of the semantic artefacts is scheduled to be performed every night.

This integration allows users to select the value of the attributes for the metadata schema directly using semantic artefacts published within EcoPortal, thereby facilitating the autocompletion.

The interaction with EcoPortal consists in searching for concepts/classes within the editing interface and subsequently retrieving information such as label, definition, URI, etc. This process, implemented for both the Data Portal and the Metadata Catalogue, facilitates the unambiguous annotation of metadata with concepts and classes from FAIR semantic artefacts.

Figure 1 shows the workflow of semantic annotation within the Data Portal, in particular for the “Keywords” attribute, which is followed for other attributes like “Software” and “Protocol”. Users are able to access (Fig. 1a, b) and search (Fig. 1c) concepts and classes from semantic artefacts published within EcoPortal through a wizard. Users can select concepts and classes that are then reported, with their URI, inside the “Keywords” element (Fig. 1d).

A similar approach is followed for the attributes of the data table but it differs from the previous one because the definition of concepts or classes is also retrieved and is reported inside the element set as shown in Figure 2.

 

 

 

 

 

 

• Make ecological (meta)data discovery more easy;
• Make annotation of ecological (meta)data easy and quick  to perform for users;
• Make meta(data) machine-actionable and machine-interpretable;
• Push towards Linked Open Data