Knowledge Management Glossary

KM Pro

Glossary Term:

Definition:

    Categorization is the process of dividing up a particular part of the world (or the whole universe) into divisions that are as mutually exclusive as possible, and possibly can be divided into hierarchical structures. These divisions are characterized as concepts. The process can be called conceptualization. The concept names provide us with a vocabulary for discussing the world.

    Categorization began in earnest with the work of Aristotle. Plato had coined the new term "Ideas," which were concepts in a non-material realm that controlled their poor copies in the material world. The Platonic Idea of a circle was perfect. Material circles only imperfect approximations.

    Aristotle criticized his teacher, and argued that his Ideas were merely generalizations - abstract concepts arrived at by finding common elements in many similar material things. Thus the ideal circle was an unachievable perfect version of all the real circles we perceive in the world. He argued that the ontological existence of Ideas was merely in human minds.

    When the concepts are shared among a CommunityOfDiscourse?, it makes intelligent discussion possible for the part of the world described. Without concepts, we cannot discuss, let alone understand, the world.

    An important categorization is the division of the natural sciences into disciplines - physics, astronomy, biology, etc. - by Francis Bacon in the early seventeenth century.

    An Ontology is a shared conceptualization. In Philosophy?, there is an implied commitment to the existence of the concepts. In Artificial Intelligence, the concepts need to be "formal," that is machine readable. In Linguistics?, "shared" emphasizes the consensual nature of concepts (as words in some human language), merely conventions of the CommunityOfDiscourse?. The part of the world described is called the "domain." A Web Ontology Language (OWL?) is a machine-processable set of semantics that allows the descriptions to be communicated between humans and their agents, application programs.

    The Semantic Web is an Ontology that uses the Resource Description Framework to define all objects (resources) by means of statements about their relations to other resources.

    Aristotle's original categories are still the rough basis for modern library categories (like the DeweyDecimalSystem).

    Categories are often expressed as a Taxonomy, emphasizing hierarchical relationships (BT and NT), or a Thesaurus, including lateral relationships (RT).

    Auto Categorization programs attempt to create categories for a collection of content objects by analyzing their text and finding common concepts to describe and sort them.

    Classification is a related activity, the sorting into predefined classes (perhaps categories) of an assortment of objects (e.g., library books).

    Related Terms: Classification, Taxonomy, Ontology, CommunityOfDiscourse?, Faceted Classification, Semantic Web

    Knowledge Management Glossary Index | Back