The subject of this master's thesis was to find all of the grammatical forms for given lemmas, extract all lexical n-grams which contain derived forms from croatian n-gram model, and, in the end, convert lexical n-grams into grammatical n-grams while paying close attention to correct classification of homographs based on context of their occurrence.
Osveta je opasna stvar, prijateljuParaCrawl Corpus ParaCrawl Corpus