Ontologies

"People can't share knowledge if they don't speak a common language."
(Tom Davenport, Lawrence Prusak: Working Knowledge)

In the field of knowledge management ontologies provide a certain terminology for the exchange of knowledge between different IT-systems and between systems and users. Ontologies are representations (or models) of relevant parts of a certain subject matter, which don’t serve the needs of a particular application, but are to be used from different users and applications likewise.

In order to built ontologies, a subject matter needs to be formalized to ensure that the used concepts are valid and understandable (precision of definition). In addition, formalized models may function as prerequisites to provide more elaborate services.

Because of its relevance to people and various applications, the knowledge contained in ontologies is quite valuable. Since – on the other hand –formalization may require enormous efforts - it is essential to choose the appropriate level in every single instance (from specific terms, taxonomies, semantic webs to complex web-languages like OWL).

Research at the Knowledge Management Laboratory attends to the following topics:

• ontology infrastructure: languages for representation, distributed ontology-services based on semantic web technologies and agents

• creating ontologies: methods and tools for building and maintaining  ontologies. The activation of knowledge stored in individual workspace computers to maintain organizational ontologies

• integration of ontologies: the semi-automatic detection of similarities in different ontologies

• ontologies in knowledge management systems

  - to support visualization and to structure portals

  - as background knowledge for searching documents and other objects

  - to identify personal or task required information needs

  - as base for integrating information from heterogeneous sources

• The evaluation from ontology based knowledge management systems

Our systems and prototypes are characterized by ontologies being combined with models of dynamic behavior (business processes, processes of knowledge work) in order to enable the personalized and task specific activation of knowledge and information.

Ansprechpartner
Dipl.-Inform. Ludger van Elst