Extending wordnets by learning from multiple resources

In this paper we present an automatic, language-independent approach to extend an existing wordnet by recycling existing freely available bilingual resources, such as machine-readable dictionaries and on-line encyclopaedias. The approach is applied to Slovene and French. The words extracted from the bilingual resources are assigned one or several synset ids based on a classifier that relies on several features, including distributional similarity. Automatic and manual evaluation shows that the resulting extensions of sloWNet and WOLF are lexico-semantic repositories of high coverage as well as high quality.