Metadaten

Ganz allgemein trifft das meiste, was oben bereits über Daten gesagt wurde, ebenso auf Metadaten zu. Informell betrachtet sind die Karteikartensysteme der Bibliothekare die ersten standardisierten Metadaten. In Bibliotheken werden Titel, Autor, Editor, ISBN etc. in strukturgleicher Form archiviert, um die Orientierung in der ständig wachsenden Anzahl der Publikationen zu erleichtern.
Tim Berners-Lee definiert Metadaten für seine Vision des Semantic Web:

Metadata is machine understandable information about web resources or other things. [Berners-Lee, 1997]

Weiter stellt er fünf Axiome für Metadaten auf:

1. Metadata is data.
2. Metadata may refer to any resource which has a URI. ³
3. Metadata may be stored in any resource no matter to which resource it refers.
4. Metadata can be regarded as a set of assertions, each assertion being about a resource (A u1 ...). Assertions which state a named relationship between two resources are known links (A u1 u2). Assertion types (including link relationships) should be first class objects in the sense that they should be able to be defined in addressable resources and referred to by the address of that resource A in { u }
5. The development of new assertion types and link relationships should be done in a consistent manner so that these sort of assertions can be treated generically by people and by software.