Answers are grounded in this paper and cite their sources.
The increasing adoption of digital technologies in mining generates large volumes of operational data that can support intelligent monitoring and decision‐making. However, accurately determining whether a gold mining site is active or inactive remains challenging because mining operations involve multiple interrelated operational, workforce, energy, and production factors. This study proposes the Artificial Intelligence‐Based Smart Gold Mining Site Status Classification Using Deep Learning (AISME‐DLOAR) framework for automated classification of gold mining sites into Active and Inactive categories. The proposed framework consists of data preprocessing, hybrid wrapper‐based feature selection using Boruta and Recursive Feature Elimination (RFE), and a Graph Convolutional Neural Network integrated with Multi‐Head Attention (GCNN‐MHA). In the graph representation, each mining‐site record is treated as a node, while similarity relationships among mining sites are represented as edges, enabling the model to learn both feature characteristics and relationships among related sites. AdamW optimization is employed to improve training stability and generalization. Experiments on the Traditional Gold Mining in Sudan dataset containing 22,398 records demonstrate that the proposed framework achieves 98.52% accuracy, 98.60% precision, 98.52% recall, 98.56% F1‐score, and 97.12% MCC on the 70:30 testing configuration. These results demonstrate the effectiveness of AISME‐DLOAR for intelligent gold mining‐site status classification and operational monitoring.
No takes yet. Share an insight, caveat, or question.
Alymani et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: