Resumen
Nowadays, data sets are spreading continually, generated by different devices and systems. The modern GPS based tracking systems and the electronic tickets are producing lots of data, and we could use them, for improving the service level. These data are processable with the modern devices and methods, and we can use them for obtaining information. Thanks to the spread of data mining, these tools are not appearing only in marketing research, but also in the most various kind of scientific areas and they are advertising a new scientific revolution. Although the importance of these data sources is essential it is not widespread in transport planning except in some specific areas. The smart card systems store the number of boarding passengers and in some cases also the alighting values. From the passengers? boarding and alighting information in a stop point we can create a time series, which shows the behavior type of the given stop points presented on graphic curves. With the help of different clustering and classification processes, these curves can be turned into groups and we can observe these groups of stop points which are defining separated zones. This is the basic step in transport modelling and the zones were determined by manual methods usually. In this paper we examine clustering and classification methods compared to each other and check the usability of different distance measurement techniques. This paper shows the usage of these methods in public transportation and presents the background of this kind of zone distribution technic.