OntologyRoadmap

This is a development area for a one-page outline of a roadmap for advancing a community ontology. To quote Donald Hobern - "A one-page outline of requirements for ontology components to support biodiversity informatics is required for inclusion in the final e-Biosphere report and TDWG has been given the responsibility for developing it."

Please (please) feel very free to contribute.

• Establish standards and/or best practicies for identifier syntax -- BobMorris - 10 Aug 2009
• Establish a clear distinction between vocabularies, thesauri and ontologies. The former are what most domain scientists care most about, the latter are what most knowledge modelers care most about. Use SKOS (now a W3 Candidate Recommendation) for thesauri and OWL for ontologies, preferably OWL DL. -- BobMorris and LeeBelbin - 28 Aug 2009
• We need to include basic statements that provide a context for an ontology otherwise the audience will not grasp the significance of the priority. I'd suggest something like this-

An ontology is a fundamental component for anyone wanting to share biodiversity data. An ontology is a formalized set of terms that fully describe a domain of interest. Not only are the terms defined precisely, the relationships between the terms is also defined precisely and formally. This formality or structural integrity enables reasoning across the ontology. An ontology unifies a domain, avoids redundancy, and links to other ontologies of adjacent or overlapping domains. -- LeeBelbin 11 Aug 2009

Vocabulary Evolution

The TDWG way of delegating standards development to small working groups with a special interest or need or desire for endorsement naturally results in standards that are often specific to a single class of applications. Many of these standards are functionally closed systems making little reference to external well known vocabularies or even existing TDWG standards. The biodiversity informatics community, expecting openness and reusability, is left to compromise real world requirements against a choice of "standards".

This approach disenfranchises many TDWG constituents and now runs counter to our obligation to bring biodiversity informatics to the semantic web.

Today, TDWG's aims of collaboration, dissemination and interchange should require that we embrace the Semantic Web by developing standards facilitating publication of Linked Data and the integration of biodiversity information within the web at large. To achieve this in a way that allows inference within the biodiversity informatics space we need to work together to develop application independent vocabularies comprising terms with unambiguous meaning.

Roger Hyam, with some help from the TDWG technical architecture group (TAG), has seeded the ground with TDWG vocabularies and these are already being used (Puma concolor species page, Sapindaceae ), together with the TDWG XML standards, to publish biodiversity content as linked data. Online feedback suggests that some change is desirable, either in the way we are trying to use these vocabularies or in the vocabularies themselves. It is also obvious that there are many within our community with the knowledge, experience, desire and skill to contribute to the evolution of a TDWG ontology if there were ways and means of doing so.

There is a real requirement for the TDWG ontology to be opened for collaborative development. Opened in a way that allows for simple wiki based browsing and contribution from registered users; including the lookup and submission of terms and definitions using application services.

As Bob suggests vocabulary evolution (SKOS?) should run in parallel with a similar process for development of a more comprehensive TDWG ontology ( OWL DL?).

-- GregWhitbread - 11 Aug 2009

OntologyServices

OntologyTools (for investigation, additions, ... comments. There are so many of these ...)

• OntologyServices
• OntologyTools
• WebHome
• WebStatistics