It is shown that basic language processes such as the production of free word associations and the generation of synonyms can be simulated using statistical models that analyze th...
The paper describes ongoing work on the evaluation of methods for extracting collocation candidates from large text corpora. Our research is based on a German treebank corpus used...
We apply pattern-based methods for collecting hypernym relations from the web. We compare our approach with hypernym extraction from morphological clues and from large text corpor...