Blockchain as a Trajectory in a Graph Database: A Model for Ethereum Transaction Analytics

Blockchain platforms generate large volumes of interconnected transactional data, whose analysis can reveal important patterns of account activity, user interaction, and value transfers. This paper proposes a graph database model for representing and analyzing Ethereum blockchain data. The proposed model captures accounts, blocks, and external and internal transactions as graph nodes, while relationships represent the decomposition of external transactions into internal transactions, sender and receiver participation, temporal block succession, and transaction recording. By interpreting the blockchain as a trajectory of consecutive blocks, the model enables trajectory-style analytics over selected temporal windows. The model was implemented in Neo4j 2026.01.3, using Ethereum data collected through the Etherscan API, and evaluated through Cypher queries over cumulative block ranges of 10, 50, 100, and 1000 blocks. The queries analyze account activity, sender–receiver relationships, statistical properties of external and internal transactions, and transaction volumes. The experimental results show that the proposed graph representation can effectively identify highly active accounts, key intermediaries such as the Wrapped Ether contract, recurrent sender–receiver pairs, and stable behavioral patterns across different block ranges. The findings demonstrate the suitability of graph databases for blockchain transaction analysis and provide a foundation for future work on larger-scale blockchain analytics, anomaly detection, community discovery, and real-time graph-based monitoring.