Impact of Data Quality on Criminal-Network Intervention
An empirical analysis of how incomplete and noisy network measurements degrade the effectiveness of network-science interventions that dismantle criminal organizations by removing central actors. Across graph-theoretic and machine-learning attack strategies, decentralization emerges as the dominant source of robustness: under missing-data conditions, targeted attacks on decentralized networks fail, and even centralized networks can be hardened cheaply with simple heuristics, arguing for caution in applying centrality-based targeting and for data-quality-robust network inference.
Garbage in Garbage out: Impacts of data quality on criminal network intervention Wang Ngai Yeung1
Network science models a criminal organization (e.g. a human-trafficking or drug ring) as a graph of actors and their ties, and designs interventions for law enforcement by identifying key players wi…