Learning Network Dismantling without Handcrafted Inputs

2025-08-01

AI Summary

This work addresses the NP-hard network dismantling (critical node identification) problem by proposing MIND, an end-to-end learning framework that eliminates reliance on handcrafted structural features. Methodologically, MIND leverages a graph neural network architecture integrating attention mechanisms with multi-round message-passing for hierarchical feature extraction, and introduces message-iteration profile analysis to enhance structural awareness. Training is performed exclusively on diverse, small-scale synthetic networks generated via a lightweight network synthesis strategy. Empirically, MIND achieves superior generalization and cross-network transferability, outperforming state-of-the-art methods on million-node real-world networks. It offers efficient training, strong adaptability across heterogeneous network topologies, and scalability to large-scale complex systems. Overall, MIND establishes a novel, feature-agnostic, and scalable paradigm for critical node identification in massive networks.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

The application of message-passing Graph Neural Networks has been a breakthrough for important network science problems. However, the competitive performance often relies on using handcrafted structural features as inputs, which increases computational cost and introduces bias into the otherwise purely data-driven network representations. Here, we eliminate the need for handcrafted features by introducing an attention mechanism and utilizing message-iteration profiles, in addition to an effective algorithmic approach to generate a structurally diverse training set of small synthetic networks. Thereby, we build an expressive message-passing framework and use it to efficiently solve the NP-hard problem of Network Dismantling, virtually equivalent to vital node identification, with significant real-world applications. Trained solely on diversified synthetic networks, our proposed model -- MIND: Message Iteration Network Dismantler -- generalizes to large, unseen real networks with millions of nodes, outperforming state-of-the-art network dismantling methods. Increased efficiency and generalizability of the proposed model can be leveraged beyond dismantling in a range of complex network problems.

Problem

• Eliminates need for handcrafted features in network dismantling
• Solves NP-hard network dismantling with message-passing framework
• Generalizes to large real networks, outperforming state-of-the-art methods

Innovation

• Uses attention mechanism for feature extraction
• Employs message-iteration profiles for learning
• Generates diverse synthetic training networks