Paradigmatic and Syntagmatic Relations in Knowledge Organization Systems

Most methods used in knowledge organization systems (KOS) organize knowledge in form of concepts and the relations between them, though the level of differentiating relations varies. Major knowledge representation methods make some of the relations explicit, but they use only a small number of paradigmatic semantic relations so far. This approach is somewhat contrary to the detailed and highly-professional theoretical reflections on relationships that are currently available from fields such as information science, knowledge engineering, linguistics, artificial intelligence studies and related disciplines, e.g. (Green et al. 2002), (Löbner 2002), (Khoo & Na 2006), (Storey, 1993), (Sun & Zheng 2005). Our aim is to point out some recent developments in knowledge representation and indexing, which motivate to reconsider the classification of relations for practical use. First of all, the current efforts to establish a Semantic Web demand for more sophisticated depiction of knowledge relationships. Ontologies, the core technique for indexing documents with semantic representations, can make use of a wider range of relations than classical KOS. A closer look at the expressiveness of logic-based ontology languages and an examination of current ontologies will help to redefine a classification of relationship types. The second and less obvious stimulus lies in upcoming folksonomies and the relations implicitly existent between user-created tags in social tagging systems (Peters & Stock 2007; 2008). Both new methods of knowledge representation may be a valuable resource for reconsidering existing relations and establishing new generalizable ones. Finally, we point out several aspects in establishing knowledge-relationships, which are of importance for information structuring and retrieval and should become subjects of future discussions.