Knowledge Organization System (KOS) for biodiversity information resources, GBIF KOS work program




The Biodiversity Information Standards (TDWG) maintains standards for biodiversity data. Many of these standards have in the past been expressed using the XML schema language (XSD). With the advance of the semantic web there is a growing interest in TDWG for expressing vocabularies as RDF resources. The Global Biodiversity Information Facility (GBIF) has proposed a Vocabulary Management Task Group (VoMaG) [1] to develop best practices and guidelines for maintaining RDF vocabularies of terms and concepts from biodiversity informatics. One of the first tasks for the vocabulary management task group would be to evaluate software tools including the ISOcat [2] and the Semantic MediaWiki [3] for collaborative development and maintenance of vocabularies of fundamental concepts (declared here to be re-used by other resources). The simple knowledge organization system (SKOS) has been proposed for the declaration of the “concept vocabularies”. SKOS vocabularies can be expressed using the resource description framework (RDF). One of the uses for the concept vocabularies is as a repository of terms for the data sharing profiles in use by the GBIF network. These data sharing profiles include the Darwin Core "extensions" and the "vocabularies" of controlled values that are declared for some of the terms included in "extensions". The overall outline is that terms to be included in the "extensions" and in the "vocabularies" would be drawn from the fundamental concepts declared by the concept vocabularies. We have proposed as a best practice to separate the declaration of standalone concepts from the declaration of semantic relationships between these concepts. The rationale is partly that the user-friendly declaration of standalone concepts will maximize their reuse, and in part that we believe the ontologies declaring semantic relationships between concepts will always depend on the purpose and context of the ontology. We believe that declaring fundamental concepts as part of rich ontologies will cause an undesired limitation for the reuse of these concepts. The web ontology language (OWL) has been proposed for the declaration of relationships between concepts. OWL can be expressed using RDF. GBIF has published a resources repository [4] for biodiversity vocabularies. The NCBO BioPortal has been proposed as a more appropriate portal for the publication of biodiversity ontologies and vocabulary resources. The BioPortal provides tools for cross-mapping between vocabulary and ontology concepts. BioPortal will therefore provide an efficient platform to build interoperability between biodiversity information standards and the standards of the biomedical community. The Darwin Core standard of the biodiversity community was loaded to the BioPortal as a first test [5].






[5] Darwin Core



Dag Endresen is the Node manager for the Norwegian participant node of the Global Biodiversity Information Facility (GBIF). Dag was previously the Knowledge System Engineer at the GBIF secretariat between August 2001 and September 2012. As a member of the research community for plant genetic resources for food and agriculture, Dag contributed to the interoperability of data standards for biodiversity information resources. He has also contributed to the development and technical implementation of data sharing protocols and software tools. Dag holds a Ph.D. in agriculture and ecology from the University of Copenhagen with the focus on prediction modeling targeting the identification of genotypes with desired trait properties.

Éamonn Ó Tuama is the senior programme officer for Inventory, Discovery, and Access (IDA) at the Global Biodiversity Information Facility (GBIF) secretariat in Copenhagen. Éamonn is responsible for data integration and interoperability issues for biodiversity data at GBIF. He has a background in marine science, holding a Ph.D. in zoology from the National University of Ireland, and several years experience in teaching and research.