Vzorkování lokálních struktur rozsáhlých síťových dat

Abstract

This Master’s thesis describes graph theory, attributes of network, sampling alghoritms and a back-in-time sampling method. This thesis is dedicated to the temporal networks and provides examples of real networks and describes real datasets that were used in experiments. The main goals of this thesis are experiments providing information about the best sampling alghoritm. For the purpose of experiments, generated models and real datasets are used. The best sampling method is evaluated by comparison of sample quality. Sampling alghoritms were selected to cover all basic principles, but most of them belong to a group based on random walk.