Generalized network dismantling based on cost-aware source-sampling betweenness

Network dismantling is critical for applications, such as infrastructure protection and epidemic control; yet, existing methods often lack efficiency and cost-awareness under real-world constraints. We propose the Cost-Aware Source-Sampling Betweenness (CASS-Bet) algorithm, a scalable framework that dynamically balances node importance and removal costs to optimize network disruption. By dynamically prioritizing critical nodes based on real-time network changes and employing a scalable sampling technique, CASS-Bet maintains high computational efficiency while enabling flexible cost definitions tailored to practical scenarios. Extensive experiments across social, infrastructure, and criminal networks demonstrate its superiority over state-of-the-art methods, enabling cost-effective dismantling with minimal resource expenditure. The algorithm's flexibility and scalability make it a practical solution for real-world challenges, from enhancing infrastructure resilience to disrupting organized crime networks.