classification

from Wikipedia, the free encyclopedia
Example of a mono-hierarchical classification
Example of a classification of a two-dimensional feature space in 5 classes and classification of an object

A classification , typification or system (from the Greek adjective συστηματική [τέχνη], systēmatikē [technē] "the systematic [procedure]") is a systematic collection of abstract classes (including concepts , types or categories ) that are used for delimitation and order . The individual classes are usually obtained by means of classification - that is, by classifying objects on the basis of certain matching features . Numerous classifications are structured hierarchically in levels with different differentiations. The set of class names form a controlled vocabulary . The application of a classification to an object by selecting a suitable class from the given classification is called classification or class assignment .

Classifications are used, among other things, in the form of taxonomies or typologies in the various sciences. Examples of systematics are the biological systematics , based on the Systema Naturae by Carl von Linné , the International Classification of Diseases ( ICD ) and various library systems.

The basic cognitive ability to form classes is known as categorical thinking , because human concept formation is already based on class formation for any objects or events of everyday perception.

Taxonomy and typology

Although the terms “taxonomy” and “typology” are often used synonymously, there are clear differences between taxonomic and typological classification systems.

Taxonomy

Scientific disciplines tend to use the term “taxonomy” (also “natural classification”), which usually stands for a hierarchical system (classes, subclasses, etc.) that

  • empirical (based on repeatable experience),
  • inductive ( inferring from the specific to the general) and
  • quantitative (large number of compared features)

was created. The classic model of taxonomy comes from biology. Therefore, such systems are often (but not necessarily) structured hierarchically and depict homologous evolutionary processes. It is classified according to origin, formation or relationship (genetic classification).

Typology

The “typology” (also known as “artificial classification”) is often used predominantly in social science contexts, the

  • conceptual (based on synthetic class formation),
  • deductive ( inferring from the general to the particular) and
  • qualitative (specially selected features)

is derived. For typologies (including the outdated racial theories of humans, ideological or psychological type theories , the concept of cultural areas or the currently existing systematics for ethnic religions), a few “typical” characteristics are used that constitute a class. The decisive factor here is often less the actual relationship, but rather a phenotypic or phenomenological classification according to analogous properties, which have a similar appearance but do not necessarily have to be related. In addition, these features are never equally pronounced for all objects of a class, but rather they range from the “ideal type” to weakly pronounced marginal types, so that clear demarcations are usually not possible. In most cases there are also objects whose assignment is controversial or even impossible.

Methodology and structure

Classifications usually include individual information objects according to a simple document object model with a strictly hierarchical structure. Therefore, neither final nor causal nor temporal chains are recorded.

Mono- and poly-hierarchical systematics

In principle, two classification structures can be distinguished: In a monohierarchy (also called a strong hierarchy or hierarchy with single inheritance), each class has only one superclass, so that the entire classification has a tree structure . In the case of polyhierarchy (called a weak hierarchy or hierarchy with multiple inheritance), a class can also be subordinated to several superclasses. If the polyhierarchy is more pronounced and there are additional relationships between the classes, it is more likely to be referred to as a thesaurus . In biology, too, one speaks of systematics when assigning species.

Analytical and synthetic classification

Another distinction is made between “analytical classification” (from general to particular, geared towards pre - coordination ) and “synthetic classification” (from particular to general, geared towards post - coordination ). Many classifications are more analytical; a prominent example of synthetic classification is facet classification .

Classification of terms

If, in the classification of terms, the items falling under term (A) also fall under term (B), then this establishes an order between the two terms (A) and (B). Every object of the subordinate concept (A) is at the same time an object of the superordinate concept (B). One then speaks of an “is-a” or “is-a” relationship between the terms (A) and (B). Example: The term electric motor is a subordinate term of the engine and therefore inherits its properties, for example the fact that it is a machine.

Other classifications can be made, for example, according to the following relationships: “is part of” ( mereology ), “is a member of”, “is created by”. With these classifications there is no inheritance of properties of the objects.

In classification systems, two types of designation can be distinguished for the terms or classes:

  • Verbal naming of terms from natural language
  • Artificial names using a notation that can consist of numbers, special characters or letters. The objects stored in a classification can be identified by means of a signature .

commitment

Systematics are used for documentation (there one speaks more of "classification"), in document management (there in connection with indexing with metadata ), in merchandise management (there one speaks more of "product groups") and in science (there one speaks more of from "systematics"). The aim of a system is to provide an overview of the objects arranged in it (analysis) and to enable thematic searches among them (order).

The services of classification systems are:

  • Summary of isolated content into classes,
  • clearer description of terms through notations,
  • Avoidance of apparent family relationships,
  • improved precision and avoidance of ballast when retrieving information .

The advantages of classification systems are:

  • Universality, i.e. orientation to the entire field of science (universal classification) or to sub-areas (subject classifications),
  • Continuity, i.e. use over a longer period of time,
  • Topicality, i.e. the ability to take new findings into account,
  • Flexibility through expansiveness (i.e. the possibility of expanding the classification system),
  • good applicability in the context of the World Wide Web, since classification systems can be mapped well as hypertext systems ( e.g. Open Directory Project ), whereby other opposing concepts also perform well in this context ( e.g. WebSom: Self-Organizing-Map ).

Disadvantages of classification systems are:

  • The system is fixed and relatively immobile,
  • Often it is hardly possible to define such a system in advance.
  • predominantly hierarchical structures,
  • no syntagmatic connection of the terms,
  • an adaptation to the progress of the subject is usually difficult to implement,
  • Issues are often "forced" into classes into which they do not fully fit, which can make the search process more difficult and lead to a possible loss of information.
  • mostly residual objects arise that do not fit into any of the listed classes and thus require a theoretically unsatisfactory residual category ,
  • mostly no objective criteria when sorting new entries: it is not always clear which category an entry is in,
  • only one way leads to the searched category (in contrast to a network-like arrangement of topic groups).

Example of the classification of a book

In the Regensburg composite classification there is the class with the notation NU 3025for the history of the Humboldt University of Berlin . The associated classification is as follows:

  • N story
  • NU History of Science and Education
  • NU 1500-7950 History of the Sciences
  • NU 2500-4250 History of Scientific Institutions
  • NU 2500-4215 Universities and Colleges
  • NU 3000-3329 German-speaking universities
  • NU 3025 Berlin / Humboldt University

Most classifications are strictly mono-hierarchical , that is, a class can only have one superclass. To clarify the meaning of individual classes, comments (so-called scope notes ) and references between related classes are used. In most systems, objects can also be assigned to several classes.

The book fellow students from 1933 on the expulsion of students from Berlin's Humboldt University is also assigned to the classes AL 50712(History of higher education and university systems at Humboldt University) and NU 7100(Other history of students as part of the history of science). In other cases, however, a class must suffice as a classification. The signature in libraries, which designates the location of an individual book, must be unique, as it can only be placed in one place. Conversely, however, several books can have the same signature.

Examples

literature

  • Jutta Bertram: Introduction to the content indexing. Basics - methods - instruments. ERGON Verlag, Würzburg 2005.
  • Traugott Koch et al .: The role of classification schemes in Internet resource description and discovery ( Memento of September 27, 2007 in the Internet Archive ) Publication of the DESIRE project, approx. 1997. Detailed overview of existing systematics (with regard to the classification of Internet content).
  • Konrad Umlauf: Introduction to library classification theory and practice. Berlin handouts for library and information science, with exercises. Berlin 1999, ( ib.hu-berlin.de ).

Web links

Wiktionary: Classification  - explanations of meanings, word origins, synonyms, translations

Individual evidence

  1. a b Daniel Hasler: Business models of the data industry: Deriving a classification approach with case studies from telematics . Diplomica, Hamburg 2014, ISBN 978-3-95850-814-9 , pp. 24 .
  2. ^ A b Christiane Hipp: Innovation processes in the service sector: A theoretically and empirically based innovation typology . Springer, Berlin / Heidelberg 2013, ISBN 978-3-7908-1264-0 , pp. 116 .
  3. Christian Lehmann: 'Typology' vs. 'Classification'. University of Erfurt, accessed on November 1, 2015 .
  4. ^ Homepage of the Federal Statistical Office. Federal Statistical Office, archived from the original on February 8, 2012 ; accessed on November 1, 2015 . Classification database. Statistics Austria, accessed on November 1, 2015 .