Design and implementation of a semi-automated method of recognition and classification of gaps and mismatches between the data in plWordNet and Princeton WordNet on the basis of the notion of equivalence