Computational approaches to suspicion in adversarial settings

摘要: Intelligence and law enforcement agencies collect large datasets, but have difficulty focusing analyst attention on the most significant records structures within them. We address this problem using suspicion, which we interpret as relevant anomaly, measure associated with data individuals. For datasets collected about widespread activities in signs of adversarial activity are rare, suggest ways to build predictive models suspicion. result lawful interception, a model suspicion spreading social network implied by intercepted data.