Benford's law application to selected data from the study information system

This diploma thesis deals with the issue of Benford's law and its applicability to the data of the study information system. Conformity to Benford's law is proved by results of analysis done on several data sets and types of data. It also deals with the possibility of detecting possible attacks on this type of system. Unfortunately, the regularities of the Benford distribution can only be used to a limited extent for this purpose. Requirements for detection using Benford's law are following: a strong profile of the attack, which doesn't comply with Benford's law, and at the same time a large number of operations performed within this attack. Out of three types of attacks, which were examined, the combination of these factors is met by an advanced automated attack, which was successfully detected using Benford's law. Another type of attack was only partially detected. It has a strong profile, but a smaller number of actions - a simple automated attack. Although its identification hasn't been confirmed by statistical methods, its impact is visible in comparison with the surrounding segments. On the contrary, cyberattack doesn't meet any of these requirements and is therefore not detectable by this method.