SUMMARY
This research propose a deep analysis of Milanese real estate market, based on data supplied by three real estate organizations; gentrification appears in some neighborhoods, such as Tortona, Porta Genova, Bovisa, Isola Garibaldi: the latest is the subject of the final analysis, by surveying of physical and social state of the area. The survey takes place in two periods (2003 and 2009) to compare the evolution of gentrification. The results of surveys has been employed in a simulation by multi-agent system model, to foresee long term evolution of the phenomenon. These neighborhood micro-indicators allow to put in evidence actual trends, conditioning a local real estate market, which can translate themselves in phenomena such as gentrification. In present analysis, the employ of cellular automata models applied to a neighborhood in Milan (Isola Garibaldi) produced the dynamic simulation of gentrification trend during a very long time: the cyclical phenomenon (one loop holds a period of twenty – thirty years) appears sometimes during a theoretical time of 100 – 120 – 150 years. Simulation of long period scenarios by multi-agent systems and cellular automata provides estimator with powerful tool, without limits in implementing it, able to support him in appraisal judge. It stands also to reason that such a tool can sustain urban planning and related evaluation processes.