ABSTRACT With the widespread adoption of Industrial Internet of Things (IIoT) technologies, the automatic classification and intelligent response to English fault description texts in cross‐border operation and maintenance (O&M) scenarios have become key to improving O&M efficiency. However, existing methods often target single languages or specific domains, making it difficult to address the complex challenges posed by multilingualism, terminology variation, and text noise in cross‐border scenarios. This leads to insufficient accuracy and timeliness of responses. To address this, this paper proposes an intelligent classification and response framework for fault texts in cross‐border O&M. The framework integrates feature fusion, denoising, and reinforcement learning mechanisms to enhance understanding and response quality for complex fault texts through multilevel feature extraction and adaptive optimization. Experiments on a self‐built cross‐border O&M dataset show that the proposed method improves classification accuracy by approximately 2.2% compared to the optimal baseline (BERT), increases inference speed by about 20% (evaluated with a batch size of 32 and approx. 15 M parameters on a Tesla V100 GPU), and demonstrates stronger robustness in high‐noise environments. This study not only provides an efficient intelligent processing solution for cross‐border IIoT O&M but also offers new technical ideas for understanding and generating responses for multilingual noisy texts, making significant contributions both academically and practically.
Hongwei Wang (Thu,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: