Enhanced tokenization algorithm for personal data protection

Tokenization is one of the methods of depersonalizing personal data. This method is a bijective replacement of fragments of personal data with random elements of a certain set. One of the weaknesses of personal data protection through tokenization is the possibility of statistically assessing the probabilities of the occurrence of protected fragments of personal data. The paper proposes a method of enhancing tokenization algorithms which allows overcoming this weakness. The enhanced tokenization algorithm is slightly different in complexity from other algorithms. At the same time, the enhanced algorithm can be used both in cases of tokenization by replacing alphabets describing various fragments of personal data and in cases where personal data are divided into fragments of the same length and converted into fragments of the same length but in other alphabets.