WordNet/GermaNet Sense Clusters

Sense Alignment of GermaNet and Wiktionary

Michael Matuschek and Tristan Miller and Iryna Gurevych:

A Language-independent Sense Clustering Approach for Enhanced WSD, in: Proceedings of the 12th Konferenz zur Verarbeitung natürlicher Sprache (KONVENS 2014), p. 11-21

PDF | Conference Proceedings

Resource Download

The clusters were produced by exploiting 1:n word sense alignments between WordNet 3.0/GermaNet 8.0 and Wiktionary/OmegaWiki/Wikipedia, as produced by the Dijkstra-WSA alignment algorithm (with and without similarity-based backoff). Senses aligned to the same target sense in the other resource are in the same cluster. The files contain one cluster per line, where each cluster represents one lexeme (lemma + pos) and the senses are tab-separated.