A gentle stroll through the Species Profile Model
Robert A. Morris
Abstract
The Species Profile Model (SPM, http://wiki.tdwg.org/twiki/bin/view/SPM/WebHome) is an OWL/RDF ontology based on the TDWG Ontology (http://wiki.tdwg.org/twiki/bin/view/TAG/TDWGOntology). It was crafted by Roger Hyam with input from an intensive workshop at GBIF in 2007, documented on the SPM wiki. As is the TDWG Ontology, SPM is expressed in RDF, the Resource Description Framework, of the Worldwide Web Consortium, but we do not presume familiarity with RDF, nor will we subject the audience to any of its syntactic arcana.
An SPM object provides various named types of information about a taxon, or more precisely, about a Taxon Concept expressed in the TDWG Ontology controlled vocabulary. The associated information (SPM "InfoItems") comprise a collection of strongly typed attributes drawn currently from one of 37 classes of information about the taxonomic, ecological, and economic properties of the taxon. These include traditional morphological descriptions, information critical to the management of invasive or endangered species, and attributes important for field biology, for ecological science and for molecular studies.
SPM InfoItems, of which an SPM object may contain several, present a simple information model based on seven properties (with possibly repeated instances). One of these is redundant and ought to be deprecated. The others are of three sorts: (1) the information itself; (2) the context (e.g. temporal or geospatial) in which the information is valid); and (3) other taxa that may be associated with the modeled taxon by biological, ecological, or other relations relevant to the InfoItem. The first two types can be given as plain text, by using a controlled vocabulary, or both.
Following the easy-to-read HTML rendering due to Hyam we will start at http://rs.tdwg.org/ontology/voc/SpeciesProfileModel for a leisurely tour a through the SPM, stopping along the way for informal opinions on improvements that might be considered. For example, some of the InfoItem context information could profitably be decoupled from specimens and observations. Time permitting, we will also discuss the (mostly easily remedied) impediments to using machine reasoning on SPM objects, and why one would want to do so. A separate presentation by Catapano et al. discusses experience in generating SPM from digitized legacy systematics literature.
An SPM object provides various named types of information about a taxon, or more precisely, about a Taxon Concept expressed in the TDWG Ontology controlled vocabulary. The associated information (SPM "InfoItems") comprise a collection of strongly typed attributes drawn currently from one of 37 classes of information about the taxonomic, ecological, and economic properties of the taxon. These include traditional morphological descriptions, information critical to the management of invasive or endangered species, and attributes important for field biology, for ecological science and for molecular studies.
SPM InfoItems, of which an SPM object may contain several, present a simple information model based on seven properties (with possibly repeated instances). One of these is redundant and ought to be deprecated. The others are of three sorts: (1) the information itself; (2) the context (e.g. temporal or geospatial) in which the information is valid); and (3) other taxa that may be associated with the modeled taxon by biological, ecological, or other relations relevant to the InfoItem. The first two types can be given as plain text, by using a controlled vocabulary, or both.
Following the easy-to-read HTML rendering due to Hyam we will start at http://rs.tdwg.org/ontology/voc/SpeciesProfileModel for a leisurely tour a through the SPM, stopping along the way for informal opinions on improvements that might be considered. For example, some of the InfoItem context information could profitably be decoupled from specimens and observations. Time permitting, we will also discuss the (mostly easily remedied) impediments to using machine reasoning on SPM objects, and why one would want to do so. A separate presentation by Catapano et al. discusses experience in generating SPM from digitized legacy systematics literature.