Access to large ICT databases revolutionalize sociology: Instead of using questionnaires with limited number of target persons and biased answers we can rely on the objective email and phone call records of millions of people. The empirical observations based on these data are qualitatively new and require theoretical interpretation, which starts with constructing adequate models. In this talk I will show the first attempts to model construction. Using elementary processes of tie formation as known from social psychology as well as the natural assumption that a relationship gets stronger by use, we build a model, which reflects qualitatively the static and dynamic structure of large social networks