Research On Bitcoin Address Tracking Based On Graph Analysis

Bitcoin provides users with a certain degree of anonymity through anonymous addresses,but meantime it is used by many criminals in illegal activities such as black market transactions,which poses a great threat to social security.Therefore,it is necessary to track bitcoin addresses to combat illegal transactions related to it.At present,there have been some researches in the field of bitcoin address tracking,but their analyzed data is insufficient and effects are not very good,so it is difficult to obtain valuable information to support actual tracking and supervision.To address this challenge,this thesis is mainly based on graph analysis to research bitcoin address tracking,and visual correlation analysis of on-chain transaction information and offchain entities on the chain to obtain valuable intelligence and provide a reference for the supervision of blockchain digital currency.The main work and innovations of this thesis include the following four sections:1.Given the problem of the lack of bitcoin address entity information and the weak relationship between address and entity information,this thesis studies the bitcoin address extraction method based on graph relationship and proposes the address model of bitcoin users and the calculation method of bitcoin address similarity.The thesis obtains a large number of bitcoin addresses with entity information on the Internet and dark web,improves the efficiency of obtaining suspicious bitcoin addresses and enhanced the degree of correlation between bitcoin addresses and entity information.2.In view of the current lack of a means of visual analysis of data on-chain and off-chain of bitcoin,this thesis studies the construction of the bitcoin address information map,and proposes a user profile of the bitcoin address and an attack model based on the information map.Based on bitcoin full transaction data and large-scale entity information,the thesis constructs an information graph of bitcoin addresses and realizes the fusion analysis of information on-chain and off-chain of specific bitcoin addresses.3.In view of the inefficiency of the existing bitcoin address correlation analysis,this thesis studies the bitcoin address association analysis based on graph model,uses heuristic clustering,address weight calculation and illegal address fund flow method,further obtains the bitcoin account and entity information associated with illegal address,and analyzes the flow direction of illegal funds,which improves the efficiency and effect of correlation analysis.4.Designed and implemented a bitcoin address tracking model based on graph analysis.Through comparative experiments with other models and case analysis,it is shown that the bitcoin address tracking model based on graph analysis proposed in this thesis can effectively track and associate specific bitcoin addresses,which can meet the current data and application requirements of Bitcoin tracking and supervision,and achieve the goal of tracking the exchange address pool to which the bitcoin address belongs.In addition,the tracking model proposed in this thesis also verifies the feasibility of applying machine learning algorithms to identify illegal bitcoin addresses.The bitcoin address tracking model based on graph analysis proposed in this thesis can be applied to the tracking and forensic analysis of illegal transaction of bitcoin and other blockchain digital currencies,which is conducive to cracking down on the current illegal transactions related to digital currency on the Internet and the dark web.It can help the relevant departments to strengthen the supervision of blockchain digital currency,and also has certain reference significance for the establishment of China's sovereign digital currency supervision mechanism in the future.