18 Oct 2012

How to create a taxonomy

David Haynes delivered a one-day workshop last week on behalf of the LAICMG on how to go about organising and classifying information. The day was structured around taxonomic theory, classification and metadata, development of controlled vocabulary and a brief overview of taxonomy software tools. Practical examples and hands-on exercises throughout the day created a rounded training experience.

Source: Wikimedia / Page  837 from the
 10th edition of Linnaeus's Systema naturae (1758),
explaining the so-called "sexual system" of plants.
The term taxonomy requires clarification. Essentially, a taxonomy represents a hierarchy of items based on shared characteristics that are internal to them. A prominent example would be the Linnaean taxonomy of rank-based biological classification.

A more pertinent rendition for our purposes here would be something like ‘a systematic way of organising knowledge, providing a hierarchical structure of concepts, using terms that help in the development of a common language to aid knowledge sharing’ (Wylli in Raschen, 2005). Taxonomies organise unstructured entities (living organisms, documents, webpages, artefacts, concepts and ideas etc.) and their contents into categories with the aim of being meaningful to information seekers.

Likewise, the terms ‘classification’ and ‘ontology’ need qualifying too as they are closely related and tend to confuse. Merriam-Webster describes ‘classification’ as systematic arrangement in groups or categories according to established criteria. An example here would be the separation of animals in the classes tasty, edible and not edible. ‘Ontology’, on the other hand, is a higher order concept and branch of metaphysics concerned with the nature and relations of being or a particular theory about the nature of being or the kinds of existents. For a qualifying characterisation of the terms ontology, taxonomy and classification and their respective differences, see van Rees.

What are the benefits of creating taxonomies?
  • They promote knowledge sharing through logical classification of in-context entities
  • They aid information retrieval
  • They facilitate differentiation between the meanings of similar sounding descriptive words or words with multiple meanings
  • They structure the semantic Web
Challenges:
  • Always consider the end-user/target audience
  • Keep the taxonomy straight forward and tangible
  • Aim for solid built and scalability (future proofing)
  • Consider potential word ambiguities (use of synonyms)
The training showcased some interesting taxonomy examples, such as the Tree of Life Web Project and reflected upon common approaches to classification (DDC, faceted classification, Colon classification). The idea of metadata was discussed, including Dublin Core as a common metadata standard within the context of describing Web resources. The importance of deploying controlled languages as a means to achieving greater precision in the description of information resources and more comprehensive retrieval was also discussed in detail.

The merits of software tools for taxonomy development were also considered: automation of routine and clerical tasks, coordination of multiple contributors, keeping track of individual terms, keeping track of relationships between terms, ability to represent taxonomy/vocabulary in different formats and facilitating upkeep and maintenance of your taxonomy.

All the same, an appraisal of common software tools and specific usage examples would have greatly enhanced this part of the training.

Check out this survey on taxonomy building tools, which considers software that is readily available and variably useful for the design process.
Thesaurus tools:
MultiTes, A.K.A, Ontology Manager, TermTree, Webchoir/TCS-10
Brainstorming tools:
MindManager, Bubbl.us, FreeMind
Ontology tools:
SWOOP, SemanticWorks, Protégé, TopBraid Composer
Visualisation tools:
RDF Gravity, ManyEyes, OwlSight, Wordmap

This excellent presentation on metadata is helpful too as it discusses the different types of metadata in great detail including relevant examples.

Finally, David kindly gave me permission to share his taxonomy resource list and 7-steps to development of a classification scheme.

Ref:
Raschen, B., A resilient, evolving resource: How to create a taxonomy Business Information Review September 2005 22: 199-204, doi:10.1177/0266382105057495


0 comments:

Post a comment