The article is devoted to comparing the efficiency of algorithms for processing Bitcoin blockchain transaction database. The article describes the algorithm of vertex marking developed by the group. Based on the comparison of this and other algorithms, it is expected to identify the most effective algorithm for clustering addresses based on belonging to a single user. The Bitcoin database contains information about millions of financial transactions. Even though information about transactions is anonymous, there are methods for combining user addresses into wallets. In this article, we study algorithms of searching connectivity components, which are based on one of the methods of combining wallets based on the heuristic feature of the «total waste» of one user. The emphasis is placed on the practical aspects of implementation – hardware limitations in processing big data sets, as well as the choice of a solution for many graph connectivity components – the maximum connected set of graph vertices, in other words, a set of nonempty vertex sets and a set of vertex pairs.