Low-cost ontology development

In this paper, we present the project building new lexical resource -- shallow ontology derived from the corpora. The ontology should be used primarily for machine translation, syntactic parsing and word sense disambiguation. Currently, the ontology for Czech language is developed, but the methodology and tools are suitable for other languages with similar structure. Ontology is based on BushBank corpus, which improves handling of ambiguity in natural language. BushBank data and tools are application-driven, thus reducing the time and costs needed to annotate the corpora and develop new lexical resources.