(eng): The paper describes the model of vague verbal data dealt already in some previous papers, and completes it by new results. The idea of effective processing of verbal data is based on the principle that each vague verbal quantity has two components, quantitative and semantic one, which can be processed separately.