A DATA-DRIVEN POLICY OPTIMIZATION ANALYSIS FRAMEWORK FOR CYBERCRIME GOVERNANCE BASED ON GNN
Keywords:
Cybercrime governance, Data-driven policy, Graph neural network, Multidimensional analysis, Space-time modelingAbstract
In today' s digital age, the rapid growth of information technology has reshaped global economic, social, and cultural landscapes, yet many countries face challenges in coordinating and implementing effective cybersecurity policies. This paper first applies PCA, SEM, GWR, and LSTM to normalize multidimensional indicators of global cybercrime data, cybersecurity policy metrics, and demographic characteristics. It then builds a data-driven policy optimization framework based on graph neural networks (GNN), incorporating spatiotemporal feature modeling, policy effectiveness evaluation, and cross-border collaborative game analysis, along with multi-source data fusion and dynamic prediction. Experimental results demonstrate high prediction accuracy and policy optimization capability, providing a scientific basis for policymakers to formulate targeted cybersecurity strategies and address evolving cyber threats, thereby improving global cybersecurity levels.References
[1] Liu Jianwen. Research and Application of Network Security Situation Awareness Technology Based on Big Data. Nanchang University, 2024.
[2] Li Siyuan. Research on abnormal behavior detection method of network traffic in railway signal system. Beijing Jiaotong University, 2024.
[3] Peng Kaikang. Malicious encrypted traffic identification based on convolutional network and attention mechanism. Nanhua University, 2024.
[4] Source. Research on network intrusion detection method based on feature selection and multi-scale fusion. Xi 'an University of Science and Technology, 2023.
[5] Gao Zihan, Cheng Lu, Zhou Aiping. Internet of Things device identification method based on convolutional neural network. Computer and network, 2025, 51(04): 363-368.
[6] Cao Shunshun, Wang Shiyi, Hu Xiaoming. Research on blockchain security evaluation under network security level protection. Network security technology and application, 2025(07): 1-4.
[7] Yang Chunxia, Tao Kerui, Song Yongsheng. Ship network security state recognition based on deep learning algorithm. ship science and technology, 2023, 45(21): 193-196.
[8] Yang Jiale. A blockchain eclipse attack detection and defense method based on CNN. Nanjing University of Posts and Telecommunications, 2023.
[9] Meng Weihang. Application of Neural Network in Computer Network Security Evaluation. Metallurgy and Materials, 2023, 43(07): 157-159.