THE TRANSMISSION ENGINE: accessing biological content linked to names representing different taxonomic concepts
Marc Geoffroy & Anton Güntsch, Department of Biodiversity Informatics, Botanic Garden and Botanical Museum Berlin-Dahlem, Koenigin-Luise-Str. 6-8, 14191 Berlin, GermanyBiological content (e.g. protection status, uses, description, sequences) is linked to taxonomic concepts (represented by a taxon name with a circumscription). If relationships are defined between concepts (e.g. congruence, inclusion), then content linked to one concept can be accessed through another one. We call this a transmission of content. However, calculating the influence of the transmission on the reliability of the content data in the new context is a complex task. Abstracting, concepts are nodes and relations are edges in a graph. Content is an “attribute” of nodes and relationships are an “attribute” of edges. With this approach transmission of content amounts to “carrying” a node attribute through the graph. Rules to compute resulting relationships between nodes connected through paths must be established. Set criteria to evaluate the relevance of such paths (e. g. its maximal length) and relationships (e. g. “exclude” relationship) must be accounted for and the applicability of transmitted information has to be modified accordingly. The Transmission Engine is a tool to accomplish this task. We present a protype Java implementation of the Engine and its caching mechanism (computation within graphs is very time consuming) and sketch the computation and the representation of taxonomic concept graphs used.