Semantic artefacts are made accessible using a wide variety of mechanisms, including publication in open repositories such as Zenodo or Figshare, deployment to github, availability as a downloadable file or object in a website, or a response offered by a web service. Most of the time semantic artefacts need to be downloaded and parsed in order to have access to its content, such as concepts/ terms, relations, and metadata.
This hampers the findability of the semantic artefact and makes reuse more difficult (see Rec. 2). To solve these issues, specific repository technologies have been developed to support the publication of semantic artefacts, their content and the metadata associated with the semantic artefacts. These “semantic repositories” provide interfaces for both humans and machines to consume semantic artefacts. They are an important piece of the infrastructure underlying the implementation of FAIR principles and FAIR Semantics as pointed out in the “Turning FAIR into reality” report and action plan (European Commission Expert Group on FAIR Data, 2018).
The number of such repositories is currently increasing with domain specific repositories and registries such as Bioportal, EBI-OLS, Ecoportal, Agroportal, BODC NERC vocabulary service or more generic services such as Finto.fi, BARTOC or Research Vocabularies Australia.
In addition, several such semantic repositories have ceased operations or do not exist any longer, and raises the requirement for sustainability. Based on analogies in the data landscape, where trustworthy repositories are defined, and recommended for long-term preservation, we propose a similar approach for semantic repositories.
Such repositories should act as a trustworthy long term archive, should provide GUPRIs, publish metadata making the semantic artefact findable for humans through a dedicated user interface, and for machines through an API. Trustworthy data repositories, focused on specific disciplines, are available for data curation and preservation, and these repositories may be amenable to preservation of semantic artefacts in the short term. In the medium term, development of criteria for Trustworthy Semantic Repositories is a community responsibility, as is the establishment of mechanisms for certification of compliance. See BP-Rec. 11.
This recommendation does not aim to support any particular technology, but emphasizes the necessity to share, publish, and preserve semantic artefacts in such repositories to improve both findability and reuse over time.
Semantic artefacts are made accessible using a wide variety of mechanisms, including publication in open repositories such as Zenodo or Figshare, deployment to github, availability as a downloadable file or object in a website, or a response offered by a web service. Most of the time semantic artefacts need to be downloaded and parsed in order to have access to its content, such as concepts/ terms, relations, and metadata.
This hampers the findability of the semantic artefact and makes reuse more difficult (see Rec. 2). To solve these issues, specific repository technologies have been developed to support the publication of semantic artefacts, their content and the metadata associated with the semantic artefacts. These “semantic repositories” provide interfaces for both humans and machines to consume semantic artefacts. They are an important piece of the infrastructure underlying the implementation of FAIR principles and FAIR Semantics as pointed out in the “Turning FAIR into reality” report and action plan (European Commission Expert Group on FAIR Data, 2018).
The number of such repositories is currently increasing with domain specific repositories and registries such as Bioportal, EBI-OLS, Ecoportal, Agroportal, BODC NERC vocabulary service or more generic services such as Finto.fi, BARTOC or Research Vocabularies Australia.
In addition, several such semantic repositories have ceased operations or do not exist any longer, and raises the requirement for sustainability. Based on analogies in the data landscape, where trustworthy repositories are defined, and recommended for long-term preservation, we propose a similar approach for semantic repositories.
Such repositories should act as a trustworthy long term archive, should provide GUPRIs, publish metadata making the semantic artefact findable for humans through a dedicated user interface, and for machines through an API. Trustworthy data repositories, focused on specific disciplines, are available for data curation and preservation, and these repositories may be amenable to preservation of semantic artefacts in the short term. In the medium term, development of criteria for Trustworthy Semantic Repositories is a community responsibility, as is the establishment of mechanisms for certification of compliance. See BP-Rec. 11.
This recommendation does not aim to support any particular technology, but emphasizes the necessity to share, publish, and preserve semantic artefacts in such repositories to improve both findability and reuse over time.