| Abstract: |
The article presents a linguistic model of frequent word combinations which is adaptable for lexicographical purposes, which integrates the different approaches of idiom research, which is strictly corpus-based, and which includes contexts of word combinations in a consistent way. The module 'Frequent word combinations in German' is a central concept of the IDS project "Wissen über Wörter" (Knowledge about Words), a hypertext-based, lexical-lexicological information system, which when completed will contain approx. 300,000 entries. Statistical co-occurrence analyses here represent an important lexicographical instrument for the reconstruction of variants, semantic features and the characteristics of headworks. Moreover, in this electronic reference work frequent word combinations (collocations, idioms and other non-idiomatic expressions) are themselves given lexicographical descriptions, on the one hand in the form of co-occurrence information about each single-word entry and on the other hand in their own type of article "multiple-word entry". Finally this co-occurrence-based approach offers a sound empirical base for linguistic research. |