What is transaction graph analysis?
- Author
- CRYPTO PORT Editorial
- Published
- Updated
- Reading time
- 5 min
In short
Transaction graph analysis draws addresses as nodes and transfers as edges, then reads relationships out of the resulting structure. Because the ledger is public, anyone can build the same graph. But an edge says nothing about ownership or intent, and claims to have identified a person from one can be, and have been, wrong.
Key points
- Addresses become nodes, transfers become edges
- The data is public, so anyone can build the same graph
- An edge implies neither ownership nor intent
- Mistaken identification harms people with no involvement
Definition
An analytical method that builds a graph with addresses as vertices and transfers as edges, then infers relationships between actors from its connectivity and clusters. Conclusions remain structural inferences.
Draw the ledger as a picture and you see funds branching outward. Where one transaction bundles several inputs, those nodes are presumed to be closely related. Where one input fans out into many outputs, it reads as a distribution or a batch of payments.
Adding a time axis reveals more: an equal amount leaving for another address moments after it arrives, or the same route being taken on a schedule. A common view starts from widely known points such as exchange deposit addresses and shades the rest of the graph by how many hops away they sit.
The limits deserve to be stated plainly. An edge records only that funds moved; it proves neither shared ownership nor any relationship. Airdrops and unsolicited transfers draw edges too. Naming an individual on the strength of a graph, if wrong, inflicts serious harm on someone innocent.
Watch out for
- · Being connected and being the same person are entirely different things
- · Even an unsolicited dust payment from a stranger leaves an edge behind
- · Publicly naming someone on the basis of a graph is deeply harmful when it is wrong