| CPC G06Q 20/4016 (2013.01) [G06Q 20/065 (2013.01); G06Q 2220/00 (2013.01)] | 10 Claims |
|
1. A scam address detection method for detecting a cryptocurrency scam address using machine learning in a scam address detection apparatus, the method comprising:
acquiring information about scam addresses labeled as being used for a scam transaction and information about benign addresses labeled as being used for a normal transaction from a database;
acquiring information about a scam address group including a plurality of scam addresses determined to be owned by a same user on the basis of the information about the scam addresses;
acquiring information about a mule address group including a plurality of mule addresses used for money laundering on the basis of the scam address group;
acquiring feature information corresponding to each of the benign addresses, each of the plurality of scam addresses included in the scam address group and each of the plurality of mule addresses included in the mule address group on the basis of at least one of the information about the benign addresses, the information about the scam address group, and the information about the mule address group; and
generating a machine learning model by machine learning of the feature information corresponding to each of the benign addresses, each of the plurality of scam addressees, and each of the plurality of mule addresses, and label information corresponding to each of the benign addresses, each of the plurality of scam addressees, and each of the plurality of mule addresses,
wherein the step of acquiring the feature information comprises acquiring first feature information representing the time from the first transaction to the last transaction of a target address on the basis of the information about the benign addresses, the information about the scam address group, and the information about the mule address group,
wherein the method further comprises:
acquiring a new cryptocurrency address,
acquiring new feature information about the new cryptocurrency address,
determining whether the new cryptocurrency address is a scam address by applying the new feature information to the machine learning model,
determining, when the new cryptocurrency address directly transacts cryptocurrency with a first address included in the scam address group, a scam risk of the new cryptocurrency address as a first value, and
determining, when the new cryptocurrency address indirectly transacts the cryptocurrency with the first address included in the scam address group, the scam risk of the new cryptocurrency address as a second value being lower than the first value.
|