A model for automatic species keying.
Ecological Services, Brown & Root, Houston, TX 77001 USA 1
Ecologists are traditional clients of taxonomists for the names, categories and the very definitions of species. Each taxonomist develops a dichotomous key for identifying members of the genus or family in which he or she is interested, but these are often difficult for the ecologist to use. A framework is presented for an automatic keying system based on a database of species attributes and a keying engine that generates new steps based on responses to previous steps. The keying engine would use object-oriented programming to create keying steps, each showing the user a species characteristic ("flower color" for example) and several potential attributes ("red", "blue", etc.). The user would then examine the organism and select among the choice of attributes, which would then narrow the range of species matches and drive the next step of the key. The database would potentially contain as many attributes of as many species as can be entered, which is more of a challenge to taxonomic science than to computer science. Populating the database would create a new library of species information that could have many applications.
Keywords: Species; taxonomy; keys; computers; databases
This abstract is being presented at: 8:15 AM in session:
Oral Session #39: Theoretical Ecology.