Enriching GermaNet with verb-noun relations - a case study of lexical acquisition

Abstract

In this paper we will focus on the lexical-semantic relations in the German wordnet GermaNet. It has been shown that wordnets suffer from the relatively small number of relations between their lexical objects. It is assumed that applications in NLP and IR, in particular those relying on word sense disambiguation, can be boosted by a higher relational density of the lexical resource. We report on research and experiments in the lexical acquisition of a new type of relation from a large annotated German newspaper corpus, i.e. the relation between the verbal head of a predicate and the nominal head of its argument. We investigate how the insertion of instances of this relation into the German wordnet GermaNet affects the overall structure of the wordnet as well as the neighbourhood of the nodes which are connected by an instance of the new relation.

Resources

Details