Rule-based extraction of English verb collocates from a dependency-parsed corpus

摘要

We report on a rule-based procedure of extracting and labeling English verb collocates from a dependency-parsed corpus. Instead of relying on the syntactic labels provided by the parser, we use a simple topological sequence that we fill with the extracted collocates in a prescribed order. A more accurate syntactic labeling will be obtained from the topological fields by comparison of corresponding collocate positions across the most common syntactic alternations. So far, we have extracted and labeled verb forms and predicate complements according to their morphosyntactic structure. In the next future, we will provide the syntactic labeling of the complements.