A paper by Korean researchers describes a new way to identify ‘criminal addresses’ from the bitcoin transaction history. Published on the IEEE portal, the paper observes that the anonymous nature of bitcoin network attracts a wide range of activities, and the ability to identify illegal transactions is important for the network.
Bitcoin’s rise to fame in the mainstream has not been an easy one, as the coin, right since its early days, has been associated with illegal activities. The most infamous was the Silk Road black market online portal. The stringent KYC norms that are coming up reflect the risky nature of blockchain-run marketplaces.
The paper, titled “An evaluation of bitcoin address classification based on transaction history summarization” observes that many people who took advantage of anonymity could not be identified, and this has become a negative feature of cryptocurrencies that dissuades many governments from its adoption.
The paper has brought out new features that add to the existing ones to build a classification model. This model will detect any “abnormality” found in bitcoin network addresses. These could be many “high orders of moments of transaction time (represented by block height)”. This will summarize the transaction history efficiently. The method utilizes supervised machine learning processes.
An experimental assessment shows that the new features improve the classification of bitcoin addresses significantly.