The article is devoted to further developing the Bayesian network model (BNM) of digital images portfolio. Earlier we've build the primary portfolio cash flow map associated with specified sales groups of assets. Now, after the detailed analysis of previous data and some reassessing it based on statistical metrics used for digital images portfolios performance the primary map was restructured. This work allowed developing initial simplified BNM and further reconstructing it as the aggregate chain of interrelated nodes, each presenting mapped elementary sales group. Thus gradual reconstruction of portfolio BNM allowed us to considerably specify the degree of main risk factors influence and the overflowing process during the assets propagation from one sales group to another. We prove that mapping of suspending group as well as group of first sale assets is reasonable for understanding of risk dynamics of intellectual assets portfolios of comparable class.