In the paper, we discuss the problems which arise when modeling metadata used for describing various kinds of statistical information, including demographical one. We also implement advanced semiotic metadata model. Moreover, we enhance a well-known Generic Statistical Business Process Model by introducing new "Depersonalization" and "Providing confidentiality" processes, and also by separating search activity. It's expected that the developed models will be applied to constructing the Integrated statistical data processing system of Ukraine and the computer-aided system "Census-2012" for census data processing. Key words: semiotic metadata model, statistical business process model, depersonalization, data anonymity.
