Using Taxonomic Concepts: integrating taxonomic concept models
Kennedy, Jessie. School of Computing, Napier University, Edinburgh
Email: j.kennedy@napier.ac.ukIdentification of specimens in biological surveys involves assigning a taxonomic name to each specimen. These names refer to taxonomic concepts understood by the biologist1. In order to be clear which concept the biologist was referring when identifying (naming) the specimen it is important that the source of the definition of the concept is also recorded (although in practice this is often not the case). In order to determine if the names recorded in these data are referring to the same or “similar” concepts we require a system in which all of these concepts are defined and may be related to each other. Many taxonomic databases (with their associated models) exist, ranging from those with a single world-view of taxonomy based on names, through to those potentially supporting multiple world-views based on taxonomic concepts. As no one database currently exists which will allow us to refer to any possible concept which might be referred to in a data set in order to address the ambiguity of using taxonomic names and concepts, it is necessary to consider and resolve data from potentially all taxonomic databases. These systems have been developed for particular purposes2 and although valuable in their own right, vary in their detailed representation of taxonomic names and concepts.
This paper describes an abstract model of taxonomic concepts which can be used to gain an understanding of a particular database or dataset’s notion of taxonomic concept and can be used as the basis for the integration of data from their models. It is shown that in principal a taxonomic concept has 2 abstract properties: an identifier (name) and a definition. The identifier, depending on the type and format, can vary in its quality in terms of its ability to uniquely identify a particular taxonomic concept. The definition may consist of many possible attributes, the composition of which determine its quality in terms of the level of reasoning about the concept that is possible based on the attributes provided. As there is no agreed standard data model sufficient to deal with the particularities of all taxonomic endeavours and allow the use of taxonomic concepts and names by others, it is important that we are able to make use of concepts as currently described and held in a range of taxonomic databases. When resolving names using concepts the degree of accuracy in matching will vary depending on the purpose for which the data is intended, therefore the system will be required to resolve concept names both loosely and exactly. The model described is being investigated as a basis for the taxonomic concept resolution service in the SEEK (Scientific Environment for Ecological Knowledge) project.
1 e.g. as described in a field guide or flora
2 e.g. creating taxonomic revisions, cataloguing specimens, providing a names catalogue