Providing corpus data for a dictionary for German juridical phraseology

We report on a set of procedures to extract juridical terminology and in particular collocations and chains of collocations from German texts from the field of the protection of intellectual property. The extraction work is based on standard corpus technology, and it produces different types of output which can be made available to lexicographers: lists of term candidates, of collocation candidates, as well as example sentences which show the use of these items in context. German juridical language is a particularly interesting domain to carry out analyses of multiword expressions, as it is very rich in collocations and in chains of collocations. Many of these multiword expressions correspond to juridical concepts and thus merit being extracted and described in an electronic dictionary. We describe our sources, our tools, and examples of the output of the extraction procedures; and we discuss options for the presentation of these results towards lexicographers.