Discovering Co-Occurrence Patterns Among Blockchain Address Categories Using the FP-Growth Association Mining Algorithm

This paper focuses on identifying recurring patterns among blockchain address categories using the FP-Growth algorithm, which is known for its efficiency in mining frequent itemsets within large datasets. The study provides insights into blockchain ecosystem dynamics by analyzing category associations across different blockchain networks like Ethereum and Bitcoin. Through this analysis, significant patterns were found, such as the frequent co-occurrence of categories related to smart contracts and exchanges, highlighting the central role of these categories in blockchain interactions. Additionally, the study delves into the influence of data sources on detected patterns, revealing that various data collection methods contribute to distinct biases, which affect category associations. The findings offer practical applications for blockchain analytics, such as improving classification models, anomaly detection, and enhancing regulatory compliance. This study contributes to blockchain research by showcasing how association rule mining can improve the categorization and understanding of blockchain address behaviors. The use of FP-Growth, as opposed to more traditional methods, enables faster and more comprehensive analysis, which is particularly valuable given the extensive nature of blockchain datasets. The research also points to potential directions for future work, such as integrating temporal data to observe changes over time and exploring additional blockchain networks to broaden the scope of insights. The study emphasizes the need for continuous advancements in blockchain address analysis to support security, transparency, and regulatory initiatives within this rapidly evolving digital ecosystem.