clustering, object-oriented modeling, persistent objects, BETA

This paper concerns the problem of object-oriented perspective in numerical clustering, which brings new possibilities of combination of various strategies of clustering. We first roughly investigate the problem of numerical clustering itself. Then we introduce object model that enable to work with clusters themselves for all the time of computing, which brings new possibilities. Moreover for modeling various cluster methods we can exploit benefit of virtual procedure mechanism. The object model contains persistent objects too for it allows to work with huge amount of data not only during the time of a single application. Described approach is being tested on medical data. The whole application is developed using Mjolner BETA System and the BETA language for its rich facilities of conceptual modeling.

This paper concerns the problem of object-oriented perspective in numerical clustering, which brings new possibilities of combination of various strategies of clustering. We first roughly investigate the problem of numerical clustering itself. Then we introduce object model that enable to work with clusters themselves for all the time of computing, which brings new possibilities. Moreover for modeling various cluster methods we can exploit benefit of virtual procedure mechanism. The object model contains persistent objects too for it allows to work with huge amount of data not only during the time of a single application. Described approach is being tested on medical data. The whole application is developed using Mjolner BETA System and the BETA language for its rich facilities of conceptual modeling.