Ontology (information science)

Ontologies in computer science are mostly linguistically concise and formally ordered representations of a set of concepts and the relationships between them in a particular subject area. They are used to exchange "knowledge", in digital and formal, between application programs and services. Knowledge comprises both general knowledge and knowledge about very specific topics and events.

Ontologies contain inference rules and integrity, so rules to conclusions and to ensure their validity. Ontologies have experienced a rebirth with the idea of the Semantic Web in recent years, and are thus part of the knowledge representation in the subarea of artificial intelligence. In contrast to a taxonomy, which forms only a hierarchical subdivision that represents an ontology is a network of information with logical relations dar.

In publications mostly of a " formal explicit specification of a conceptualization " ( concept formation ) is spoken.

  • 3.1 Example Ontology
  • 3.2 ontology editors
  • 3.3 ontology
  • 6.1 Literature 6.1.1 Understanding ontology
  • 6.1.2 Biomedical Ontology
  • 6.1.3 applications


Ontologies are used as a means of structuring and data exchange to

  • Already put together existing knowledge stocks
  • To look into existing knowledge bases and edit these
  • Of types of knowledge assets to generate new instances.

The most well-known applications have no individual instances and are limited to scientific purposes to systematize the use of concept spaces. Ontologies are known for genetic data in bioinformatics or spatial information in the Geosemantik.

New applications are to be expected when the ontologies are used as types for instantiation of individual information concepts, such as in human medicine for the case-specific medical documentation, the patient record. Already developed applications in human medicine are so far provided no connection between known classification systems in clinical practice here. Instead, they bind to date merely to individual classifications for scientific work.

Experiments for profitable use of ontologies in business application software published by SAP.

In the bridge between various classifications and to neighboring worlds concept is the strength of ontological concepts: they allow the separation of the conceptual work of fixed templates and text blocks and the transition to varying combinations of semi- finished worded texts to draft individual texts.

Construction of ontologies

Analogous to a database, in the structure ( database schema ) and contents (data) form a whole part, even when an ontology together the rules and terms. While traditional databases have no information about the meaning of the stored data, ontologies have a formal description of the data and rules about their relationship. These rules make it possible to draw conclusions from the available data to identify and independently to complement the lack of knowledge of the existing contradictions in the data. These conclusions are derived by inference, that is, by logical deduction.

Under " Ontology learning" (perhaps with " ontological learning" to translate ), the process can be described, in which an ontology by automatic methods acquired further knowledge and thus increases in size and structure. For inferences play an important role. In this process knowledge is generated through an automated process, while ontologies else to gain by inputs of human experts knowledge.

From the possibility of relations on relations ( in RDF called reification ) and rules shall be made ​​relatively rare, partly due to their complexity in practice use, despite the fact that these features distinguish ontologies from other conceptual systems.


  • Terms: ( in English: "concepts ", sometimes translated with the wrong friend " concepts" ): The description of common properties is defined as a concept (eg "city" or "country" ). Terms are also referred to as classes. These may be arranged in a class structure with positive and negative range.
  • Types: types represent object types in the ontology and provide the available types in classes dar. These are generated based on pre-defined criteria and designated as type (eg city as an instance of the concept of topological element of the class points or river as an instance of concept of topological element of class lines)
  • Instances: Instances represent objects in the ontology, and make the knowledge available dar. These are generated based on pre- defined terms and referred to as individuals (eg Munich as an instance of the concept of topological location of type city or Germany as an instance of the concept topological location of type country).
  • Relations: instances of the same type must be adapted to different circumstances. These relations are used to describe the relationships between the entities exist (eg city of Munich located in Country Germany ). Relations are also referred to as properties.
  • Inheritance: It is possible to inherit properties and relations of concepts. In this case, all properties are propagated to the inheriting element. Multiple inheritance in terms is possible. By using transitivity instances can be configured in a bottom-up hierarchy. This is called delegation.
  • Axioms: axioms are statements within the ontology that are always true. These are normally used to represent knowledge that can not be derived from other terms (eg, " between America and Europe, there is no train connection. ").


Basically, one divides ontology into two types:

  • Lightweight ontologies include concepts, taxonomies and relationships between concepts and properties which describe the same.
  • Heavyweight ontologies are an extension of lightweight ontologies and add these axioms and constraints added, whereby the intended meaning of individual statements within the ontology becomes clearer.

Ontology creation

An ontology depends on by whom it is used. For example, it may be important for a restaurant with an ontology about wines to also include appropriate food to the wines in the ontology. If the user is, however, a wine bottler, the range of food is likely to be completely uninteresting. However, it is important for the bottler, which exist different cork and bottle types.

Example Ontology

The figure shows the operating principle of an ontology. The upper level shows the ontology that contains concepts and relations. Terms are represented by ellipses and relations by arrows. The rectangles represent simple container for information represents the relations linking two concepts together and limit it at the same time, for example, a work of art by an artist is generated.

Terms can be used for inheritance. For this reason, the painters and sculptors also have the relations name and first name. The thick arrow indicates the inheritance. The two relations suggests and paints and gemaltVon and geschlagenVon are inherited relations of generated and hergestelltVon. The original relations properties are retained, but may be extended.

The relations paints and gemaltVon have inverse relationships to each other, thereby further logic is integrated into the ontology, which allows that it can be of a painter on his art and vice versa, from one image to the painter closed.

The lower plane of the figure shows instances of the ontology. These are represented by a black dot. The abbreviation (I1 ) stands for the unique resource name of the instance. In the Semantic Web, a URI is used to identify. A special feature has the instance of the painter Raffaello Santi. This already uses existing instances, namely I3 of the oil drawing and I6 type Galleria dell'Accademia.

Ontology editors

→ Main article: Ontology Editor

Various software tools to support the construction of ontologies in different ontology languages.


Formal languages ​​for describing ontologies include the RDF Schema, DAML OIL, F -Logic, propagated by the World Wide Web Consortium for the semantic Web Web Ontology Language (OWL), the Web Service Modeling Language ( WSML ) and the under ISO / IEC 13250:2000 standardized Topic Maps. Also, the Knowledge Interchange Format ( KIF ) is used occasionally.


Originally ontology as the science of beings a philosophical discipline and part of metaphysics.

As a precursor of an explicit formalization of Ontologiebegriffs Charles S. Peirce and Edmund Husserl should be mentioned. A formal perspective on the philosophical ontology also had Alonzo Church in 1958, and Willard Van Orman Quine. Quine has argued a Ontologiebegriff, which broke with the tradition of the classical conception of Ontologiebegriffs in philosophy. According to Quine thinks "being": to be value of a bound variable. In the way to the truth there is the thesis: " Empirically are of concern to an ontology only said neutral nodes that it contributes to the structure of the theory. "

In the field of artificial intelligence, the term " ontology " was from the early 1990s through an article by Neches et al. and subsequent publications popular.

From then on, the term " ontology " spread out as an explicit formalization, was used in the artificial intelligence research and picked up by the bioinformatics and other subjects.

In 1999, Tim Berners -Lee in the book Weaving the Web before his vision of the Semantic Web. Many times quoted in this context, the article The Semantic Web by Berners -Lee et al in 2001, in which he also describes the use of ontologies in the context of the Semantic Web.