Account Clustering in the Polkadot Network: Heuristic, Experiments, and Insights

This paper investigates, for the first time, user account clustering in the Polkadot network, one of the most innovative account-based altcoins in the market. To achieve this goal, we levereged the 'deposit address reuse' heuristic on the Polkadot relay chain. In detail, we propose a novel deposit address detection methodology, combined with a general clustering strategy. To show the viability of our approach, we present a case study involving Binance and Kraken, the two major exchanges active in the Polkadot network. The analysis extends over a sensitive time window-starting from Polkadot genesis (May 2020) up to block 12,532,600 (October 2022). Thanks to the proposed methodology, we clustered more than 145,440 accounts belonging to exchanges, and more than 25,000 user accounts, representing around 25% of all the Binance/Kraken on-chain customers. The general applicability of our technique, the preliminary achieved results-showing both the viability and the value provided by our approach-, and the research hints discussed in the paper, also pave the way for further research in the field.