Detecting Anomalies on Cryptocurrency Markets Using Graph Algorithms

Agata Skorupka

2023

Abstract

The low level regulation of cryptocurrency market as well as crucial role of trust and digital market specificity makes it a good environment for anonymous transactions without identity verification, therefore fraudulent activities. Examples of such anomalies may be failing to fulfil transaction, as well as different forms of market manipulation. As cryptocurrencies are incorporated in more and more investment portfolios, including big companies accepting payment by this means, anomalies on cryptocurrency may pose significant systemic risk. Therefore there is a need to detect fraudulent users in a computationally efficient way. This paper presents usage of graph algorithms for that purpose. While most of the literature is focused on using structural and classical embeddings, this research proposes utilizing nodes statistics to build an accurate model with less engineering overhead as well as computational time involved.

Paper Citation

in Harvard Style

Skorupka A. (2023). Detecting Anomalies on Cryptocurrency Markets Using Graph Algorithms. In Proceedings of the 12th International Conference on Data Science, Technology and Applications - Volume 1: DATA; ISBN 978-989-758-664-4, SciTePress, pages 504-509. DOI: 10.5220/0012133900003541

in Bibtex Style

@conference{data23,
author={Agata Skorupka},
title={Detecting Anomalies on Cryptocurrency Markets Using Graph Algorithms},
booktitle={Proceedings of the 12th International Conference on Data Science, Technology and Applications - Volume 1: DATA},
year={2023},
pages={504-509},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012133900003541},
isbn={978-989-758-664-4},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 12th International Conference on Data Science, Technology and Applications - Volume 1: DATA
TI - Detecting Anomalies on Cryptocurrency Markets Using Graph Algorithms
SN - 978-989-758-664-4
AU - Skorupka A.
PY - 2023
SP - 504
EP - 509
DO - 10.5220/0012133900003541
PB - SciTePress