Automatic sense clustering in eurowordnet

This paper addresses ways in which we envisage to reduce the fine-grainedness of WordNet and express in a more systematic way the relations between its numerous sense distinctions. In the EuroWordNet project, we have distinguished various automatic methods for grouping senses into more coarse-grained sense groups. These resulting clusters reflect aspects of lexical organization, displaying a variety of semantic regularities or generalizations. In this way, the compatibility of the language-specific wordnets in the EuroWordNet multilingual knowledge base is increased. 1