An approach for combining ontology learning and semantic tagging in the ontology development process: eGovernment use case

In this paper we present a novel method for ontology development that combines ontology learning and social-tagging process. The approach is based on the idea of using tagging process as a method for refinement (pruning) of the ontology that has been learned automatically from available knowledge sources. In the nutshell of the approach is a model for the conceptual tag refinement, which basically searches for terms that are conceptually related to the tags that are assigned to an information source. In that way the meaning of the tags can be disambiguated, which support better usage of the tagging process for the ontology pruning. We have developed a software tool, an annotation framework, which realizes this idea. We present results from the first evaluation studies regarding the application of this approach in the eGovemment domain.