Representing the meanings of words is a formidable task for linguists; understanding them is arguably the greatest challenge for machines. Wordnets are lexical databases in which the meanings of words are given in terms of their relations to other words. This model of lexical meaning - inspired by research on human semantic memory - allows automatic systems designed for Natural Language Processing to detect and measure semantic similarity among words in a text.

Based on the original English WordNet, semantic networks have been constructed for donzens of genetically and typologically distinct languages, opening up new vistas on the diversity of human communication. The Proceedings of the Fifth Global WordNet conference include papers on the challenges of constructing new wordnets in a wide range of languages, the theoretical foundations and structure of wordnets as well as reports on numerous extensions and applications.