The paper presents results of research on clustering housing community properties with the use of classification tree. The subject of the research was the population of housing community properties with the Olsztyn community share. The research procedures were used in particular way that enabled acquired classification tree include all the essential criteria in terms of the rational community properties management. The classification tree presents hierarchical order of maximum homogenous subclusters with the minimum number of nods generating clusters of simple classification rules. The analysis have shown that big housing communities were dominant in the population and that it contributed to improved management efficiency. One have noticed that in the overwhelming majority of analyzed housing communities Olsztyn community lost most of the shares in a result of shattered privatization. This situation greatly complicates the decision-making process in terms of community properties management.

73-84