A Lexicographic Evaluation of German Adjective-Noun Collocations

This paper describes a small database of 1,252 German adjective-noun combinations, which have been annotated by professional lexicographers with respect to their collocational status and their usefulness for the compilation of a bilingual dictionary. The database is a random sample taken from the most frequent (f ≥ 20) adjective-noun pairs in a standard newspaper corpus (Frankfurter Rundschau). It is particularly useful for the evaluation and development of ranking techniques for multiword candidates. Suitable corpus frequency data (instances of adjective-noun cooccurrences from the same corpus) are made available together with the database.