Systems and methods for risk diagnosis of cryptocurrency addresses on blockchains using anonymous and public information

A system, comprising:

• one or more databases storing data with labels and data without labels;
• a server executing a risk diagnosis computer program;
• and
• a user device executing a user application;
• wherein:
• the risk diagnosis computer program receives the data with labels and the data without labels, fits the data without labels and the data with labels into a plurality of clusters using unsupervised learning, assigns a risk level to each of the plurality of clusters based on the data with labels in the cluster, receives unseen data from the user application, and predicts a risk level associated with the unseen data by constructing features from the unseen data, clustering the unseen data to one of the plurality of clusters, and returning the risk level associated with the cluster to which the unseen data is clustered.

A non-transitory computer readable storage medium, including instructions stored thereon, which when read and executed by one or more computer processors, cause the one or more computer processors to perform steps comprising:

• receiving data with labels and data without labels from public data databases and/or anonymous data databases, wherein the data comprises transaction graphs and public reports for blockchain addresses;
• fitting the data without labels and the data with labels into a plurality of clusters using an unsupervised learning/embedding algorithm, wherein the unsupervised learning/embedding algorithm comprises K-means clustering or a graph neural network;
• assigning a risk level to each of the plurality of clusters based on the data with labels in the cluster;
• receiving unseen data, wherein the unseen data comprises a cryptocurrency address associated with a potential counterparty;
• and
• predicting a risk level associated with the unseen data by constructing features from the unseen data, clustering the unseen data to one of the plurality of clusters, and returning the risk level associated with the cluster to which the unseen data is clustered.