|
|
Metadata, which is data on data, is used to describe data objects, e.g. the data fields
author, title and year of publication of a textual object. Problematic metadata are keywords,
because it is difficult to use them consistently (e.g.: "auto" or "car").
The solution of this problem is so-called "controlled vocabulary", which prescribes
the terms to use. Three different kinds of such systems of terms can be distinguished on the basis
of their relation types:
- Systems of classification:
One relation type: superclass - subclass
- Thesaurus:
Limited number of relation types, e.g.: synonyms (different terms with the same meaning), antonyms (terms with the opposite meaning), superclass - subclass
- Ontology:
Unlimited number of relation types, e.g.: Part - Whole relation, kinship etc.
Ontologies are also the most flexible and thus best way of storing knowledge. They are essential for the
design of a more intelligent internet (the so-called "Semantic Web", the new vision of
WWW-inventor Tim Berners-Lee).
DAASI International created a data model that maps ontologies in directory services, so that all advantages of the latter - authentication, access control, distribution in the net etc. - can be benefited from. Interfaces to existing ontology definitions (e.g. in RDF) as well as various inquiry-languages (e.g. RQL) can be implemented.
The following knowledge management solutions can be individually developed:
- Content Management systems
- E-Learning systems
- Intelligent agent-programs which receive data from portals via Web Services (SOAP, WSDL)
- Compilation of electronic university calendars
For the realization of knowledge management projects we are able to fall back on the following core competences:
- LDAP (Lightweight Directory Access Protocol)
- CIM (Common Information Model)
- RDF (Resource Description Format)
- various XML-technologies (eXtended Markup Language)
- Topic Maps
|
