Graph-Based Profiling of Blockchain Oracles

The usage of blockchain technology has been significantly expanded with smart contracts and oracles . While enables to automate the execution an agreement between untrusted parties, oracles provide data external a given blockchain, i.e., off-chain data. However, validity accuracy such can be questionable that compromises transparency immutability chacteristics blockchain. Despite many studies on trustworthiness oracles, more precisely, data, their solutions are often ‘short-sighted’ dependent binary decisions. In this paper, we present novel graph-based profiling method determine oracles. We construct graph as nodes cumulative average discrepancies edge weights. Our continues update graph, weights in particular, distinguish trustworthy Clearly. discourages provision false inaccurate have conducted evaluation study see effectiveness our proposed method, which run experiments utilizing Ethereum network. Additionally, also calculated cost running these experiments. Consequently, experiment results show achieves around 93% identifying sources.