What is it about?
As Graph Convolutional Network (GCN) has become emergent in the past few years, this method has been applied to the graph network of Bitcoin data. However, GCN is suited for non-directed graph networks. For this purpose, we have applied GCN in a novel way using a combination with linear layers and tweaking the adjacency matrix that have shown to perform better in blockchain graph network.
Featured Image
Photo by Nastya Dulhiier on Unsplash
Why is it important?
This paper highlights the competence of graph network in such dataset. We consider this paper as a great starting point in a more detailed graph networks e.g. networks accompanied with edge features.
Perspectives
Recently, GCN has become a hot topic in machine learning world. As GCN captures different patterns in the graph network, this will effectively capture some rapid patterns or vague behaviour that is done in a very complicated network! This paper has shown how GCN boosts the performance of detecting illicit activities in Bitcoin data
Ismail Alarab
Bournemouth University
Read the Original
This page is a summary of: Competence of Graph Convolutional Networks for Anti-Money Laundering in Bitcoin Blockchain, June 2020, ACM (Association for Computing Machinery),
DOI: 10.1145/3409073.3409080.
You can read the full text:
Resources
Contributors
The following have contributed to this page
