<basses:matrix.org> ART: A Graph-based Framework for Investigating Illicit Activity in Monero via Address-Ring-Transaction Structures

<basses:matrix.org> arxiv.org/abs/2511.16192

<rucknium> Nothing groundbreaking in this paper IMHO. Bases its analysis on 19 transactions from 2017.

<rucknium> Hits the same problem I have stated before: Machine learning with training sets works poorly on Monero because of the lack of valid training samples.

<rucknium> > The most significant constraint lies in the very limited support of both the positive and negative classes, reflecting the scarcity of publicly available labeled Monero data.

<rucknium> @ack-j:matrix.org may have thoughts too.

<rucknium> The grant programs that funded the paper:

<rucknium> cordis.europa.eu/project/id/101168562

<rucknium> cordis.europa.eu/project/id/101121281

<321bob321> Eu

<321bob321> The funding participants looks dodgy

<321bob321> Cyprus policeCyprusNet EU contribution€ 170 000,00

<321bob321> Random