TokenVis:面向以太坊区块链ERC-20智能合约演变模式的可视分析方法OA北大核心CSTPCD

Blockchain techniques have received extensive attention in recent years.The cryptocurrency market based on block-chain techniques is complex and unstable,vulnerable to political,economic and social factors.Existing studies fo-cus on native cryptocurrencies,such as Bitcoin and Ethereum.However,a large number of ERC-20 tokens in the cryptocur-rency market exist.ERC-20 tokens have a great market capitalization,attracting many investors'attention.This paper pro-poses TokenVis,a visual analytics system,to help investors understand the evolutionary patterns of different ERC-20 stan-dard tokens and provide explanations.The TokenVis prototype system integrates a visual analytics framework with differ-ent time granularities.We propose SegRank visualization for presenting evolutionary patterns of multiple time series and a time-based Constrained Optimization BY Linear Approximation(COBYLA)optimization algorithm to show the relation-ships between evolutionary patterns and news to provide explanations.We present two case studies involving the evolution-ary patterns of different tokens to demonstrate the effectiveness and usability of the TokenVis prototype system.