Some definitions

SKOS is a W3C recommendation designed for representation of concepts, controlled vocabularies, taxonomies, thesauri, and ontologies. It is machine-readable and perfectly suited to model knowledge-intensive systems, such as the Sensing Clues data model for Nature Conservation.


Within the scope of the Sensing Clues platform, we adhere to the following definitions.


A Concept (or topic) is a singular 'thing' that be described in many different ways.


A Controlled vocabulary is defined a simple flat list of Concepts which for one reason or another are grouped in 1 list.


A Taxonomy is an hierarchical classification system to structure and organise concepts thematically. An example of a Birds taxonomy is given below.


A thesaurus adds to a taxonomies a number of SKOS-defined attributes and relations, such as synonyms (alternative labels), references to other thesauri or taxonomies, linked open data sources (such as wikidata or geonames), and so on. It thus allows for other than hierarchical relations, providing the ingredients of a graph.


In an Ontology you can add domain and application specific properties and relations to taxonomies and thesauri. As such an ontology lays the foundation for the construction of a semantic knowledge graph.


A semantic knowledge graph is all of the above plus actual data, in our case: observations.



The Sensing Clues Data Model

To create a semantic model for the Nature Conservation domain, one that can be used across the world, the Sensing Clues database is set up as a semantic knowledge graph. It includes a custom-built ontology (SCCSSOntology), which is linked to a number of internationally renown taxonomies, such as the GBIF Backbone Taxonomy.


This linking process is far from straightforward. Taxonomic rankings are often contested, names of individual Concepts often include abbreviations, have synonyms, and may be identical to Concepts located at entirely different locations within the taxonomies. For example: there are many ways of spelling Anumbius Annumbi; common names may (wrongfully) be related to a genus, species or subspecies.


To accommodate standardised data collection and improve the quality and usability of scientific taxonomies, we've been creating taxonomies for:

  • Birds (over 11 000 species, sourced from GBIF: see GBIF Birds)
  • Mammals
  • Reptiles
  • Amphibians
  • Fish
  • Flora

The taxonomies that are (in progress of) being included in the Sensing Clues data model can be consulted with our Taxonomy Explorer.