Motivation

 

Language Technology in Ontology Development and Use

 

In recent years, the Internet evolved from a global medium for information exchange (directed mainly towards human users) into a “global, virtual work environment” (for both human users and machines). Building on the world-wide-web, developments such as grid technology, web services and the se-mantic web contributed to this transformation, the implications of which are now slowly but clearly being integrated into all areas of the new digital society (e-business, e-government, e-science, etc.) In particular, grid technology allows for distributed computing, web services for a distributed workflow, and the semantic web for increasingly intelligent and therefore autonomous processing.

In this, it is important to realize that the semantic web will function more and more as the man-machine interface of this “global, virtual work environment”. The underlying semantic web infrastructure of shared knowledge (ontologies) and markup of resources and services with such knowledge (ontology-based metadata) ensures that a common understanding will exist between the human user and the machine-based processes. However, as much of human knowledge is and will be encoded in language, multilingual and multicultural aspects (culture as specific to countries, regions and nations, connected with language) will play an important role in establishing and maintaining such common un-derstanding. Given these considerations, we emphasize the following two im-portant issues in future semantic web development:

 

Making the semantic web accessible in many languages

 

Authoring support for automatic knowledge markup should be available for many languages thereby avoiding that only documents in some languages will become part of the semantic web

 

Allowing the semantic web to represent many different cultures

 

Ontologies should express concepts as used in different cultures, thereby avoiding that the semantic web would force an unnecessary semantic stan-dardization. Therefore, tools for ontology adaptation and for mapping different ontologies should be an integral part of the semantic web infrastructure.

In both cases, there will be an important role for a combination of language technology, ontology engineering and machine learning, in order to provide text analysis for knowledge markup and text mining facilities for ontology mapping and learning. A growing integration of language technology tools into semantic web applications is therefore to be expected with the following characteristics:

 

Language Technology for the Semantic Web

 

Language technology tools will be used for efficient, (semi-)automatic knowl-edge markup (based on information extraction) and ontology development (based on text mining), allowing web documents in many languages and from different cultural backgrounds to be integrated on a large scale within the se-mantic web.

 

The Semantic Web for Language Technology

 

Semantic web methodologies (metadata, web services) and standards (RDF/S, OWL) will be used in the specification of web-based, standardized language resources – data (corpora, lexicons, grammars) and tools – allowing for a dis-tributed and widespread use of these resources in semantic web applications.