DCKH‐CNN: A Multimetric Graph‐Based Convolutional Neural Network for Identifying Key Influential Nodes in Earth Surface Data Linked Networks

• DOI: 10.1111/tgis.70016
• Publication Date: 2025-03-11T16:12:34Z
• Authors: Qinjun Qiu; Jiandong Liu; Mengqi Hao; Weijie Li; Yang Wang; Liufeng Tao; Liang Wu; Zhong Xie

ABSTRACT Identifying key influential nodes in Earth surface data association networks is crucial for optimizing the use of scientific data. However, challenges such as network size, complexity, and dynamic node influence make this task difficult. While deep learning methods have improved recognition accuracy reduced computational costs complex networks, they still struggle with balancing efficiency accuracy. To address this, we propose DCKH‐CNN, a novel Multimetric Graph‐Based Convolutional Neural Network framework. Based on LCNN model, it integrates global local features by calculating metrics degree centrality, K ‐shell, H ‐index, near‐centrality. One‐hop two‐hop adjacency matrices are used to represent internode relationships, enhancing feature representation. Trained small‐scale model captures unique characteristics. Experimental results using SIR demonstrate that DCKH‐CNN surpasses state‐of‐the‐art algorithms vast majority Surface Data Linked (ESSDLN) datasets real‐world accuracy, while demonstrating moderate time consumption. This method offers more efficient approach identifying supporting accurate recommendations intelligent analysis