Knowledge Discovery in Cryptocurrency Transactions: A Survey

Abstract

Cryptocurrencies gain trust in users by publicly disclosing the full creation and transaction history. In return, the transaction history faithfully records the whole spectrum of cryptocurrency user behaviors. This article analyzes and summarizes the existing research on knowledge discovery in the cryptocurrency transactions using data mining techniques. Specifically, we classify the existing research into three aspects, i.e., transaction tracings and blockchain address linking, the analyses of collective user behaviors, and the study of individual user behaviors. For each aspect, we present the problems, summarize the methodologies, and discuss major findings in the literature. Furthermore, an enumeration of transaction data parsing and visualization tools and services is also provided. Finally, we outline several future directions in this research area, such as the current rapid development of Decentralized Finance (De-Fi) and digital fiat money.

Community detection in complex networks: From statistical foundations to data science applications

“…In turn, the rapidly emerging blockchain technology poses a set of new challenging problems in network sciences and clustering, in particular. First, in addition to sheer amounts of information to be analyzed, blockchain transaction graphs are highly sparse and time-evolving, making application of many conventional clustering approaches infeasible (Akcora et al, 2019;Liu, Jiang, et al, 2020). Second, one of the primary goals of cluster analysis on blockchain networks is to identify addresses which are managed by the same user, rather than to group multiple users based on their similarity (Victor, 2020).…”

Characterizing key agents in the cryptocurrency economy through blockchain transaction analysis

“…Five classifiers are trained: Logistic Regression (LR), Support Vector Machine (SVM), Multilayer Perceptron (MLP), Random Forest (RF), and LightGBM. These algorithms are the most common choices and best performing ones in previous blockchain addresses identity classification tasks (mostly on Bitcoin addresses) [23]. The choices and tuning of the algorithm parameters are given in Sect.…”