Assessing the Practice of Ontology Metadata: A Survey Result
The data deluge - or Big Data - brings us to think differently about data management and shows us the urgent need to move towards FAIR (Findable, Accessible, Interoperable, Reusable) data. Semantic resources (ontologies, vocabularies, thesauri and terminologies) are no exception to this rule. Between February 16 and March 30, 2018, we conducted a survey on the use of metadata vocabularies to describe such resources. We wanted to evaluate the current state of practice and discuss recommendations in terms of metadata standards for ontologies. In this paper, we present and discuss the results of this survey. The analyze shows that the core semantic web languages (RDFS, OWL, SKOS) and vocabularies such as DCAT or Dublin Core are among the most known and used vocabularies to describe ontologies. More surprisingly, most of the numerous vocabularies really relevant for describing ontologies are barely known and never used: DOOR, VANN, ADMS, OMV, MOD. This demonstrates a lack of a clearly identified standard recommendation for this purpose. We then propose perspectives to meet the need for harmonization of metadata vocabularies, placing this survey in the context of the work already done on this question.