Improving Collocation Extraction By Using Syntactic Patterns
Ruifeng Xu, Qin Lu · 2006
A study on using syntactic patterns to improve window-based collocation extraction systems is presented. The support collocation patterns and reject collocation patterns retrieved from a chunked corpus and are used in two different strategies. The first strategy uses only the support patterns in preprocessing stage whereas the second strategy incorporates both the support and the reject patterns into the existing systems. Experimental results show that the use of syntactic patterns can significantly improve the performance of collocation extraction especially for filtering out pseudo collocations and the extraction of low-occurrence collocations.