A taxonomy is a classification scheme for organising information and data into meaningful groups. It helps agencies organise their information into hierarchical relationships to ensure it can be efficiently searched, found and its meaning correctly interpreted.
Correct and consistent understanding of shared information i.e., semantics is a fundamental theme for enabling interoperability.
In its simplest form, the hierarchical structure of taxonomy is established using parent/child relationships between objects. It uses controlled vocabulary to name each object in relation to other objects.
Taxonomy uses hierarchy and controlled vocabulary to:
- manage synonyms
- ensure consistent application of business terms
- ensure business terms are correctly interpreted
- reduce information ambiguity.
When creating a taxonomy consider using existing data definitions.
- common terms and classes across agencies
- reducing necessary data transformations
- correct interpretation of data and information.
A functions thesaurus can be used in this manner. You can use thesauri that are available for Australian Government agencies to identify new and necessary controlled vocabularies when developing or updating your taxonomy.
Taxonomies can be created using automation tools. Automatic Taxonomy Construction (ATC) is a process that uses automated tools to generate taxonomy classifications from bodies of text.
ATC tools are helpful when you are:
- needing to create large ontologies
- flagging and correcting errors in existing taxonomies
- regularly generating large amounts of data that needs to be classified.
Ontology describes an object in the same way as taxonomy – by its hierarchical position however it also describes an object by its relationships to other objects that are not in its linear hierarchy.
Eg, your family's pet cat could be represented using common language taxonomy as:
An ontology may also associate your pet with concepts of popular Youtube content or Egyptian beliefs. When comparing taxonomy and ontology, ontologies are more in-depth, complex and represent richer relationships.
Ontology languages such as Resource Description Framework (RDF) and Web Ontology Language (OWL) are commonly used to construct ontologies that reflect the rich diversity and complex data of the Semantic web. They generate machine-readable language to encode knowledge on web pages, enabling their information to be queried and analysed. These ontologies help describe classes, attributes, relations, and events and can contain a collection of taxonomies. Examples of OWL ontologies are showcased by the Australian Government Linked Data Working Group (AGLDWG).