Authors: Istvan Molnar, Gyorgy Lipovszki
The paper discusses the repetitive direct use of large data sets in microsimulation models. First, the income tax microsimulation models and the related data requirements are discussed. Next, the processing of the applied data set is introduced and the idea of using direct data instead of samples developed along with applied data stream management principles. The discussion of data stream management issues turns subsequently into an analysis of model control aspects of data driven simulation. Finally, some practical experiences and applied technologies are discussed.