This is the authors' abstract. We don't add key points for this paper.
Aiming to address the challenge of incorporating multiple fault phenomena and utilizing accompanying information such as negation and time adverbials in the process of diagnosing faults in key equipment for bearer network transmission, this study proposes a fault diagnosis method based on knowledge graph and multi-modal neural network. By constructing a knowledge graph from a fault diagnosis knowledge base, this method retrieves candidate fault root causes based on the fault phenomena and extracts subgraphs along the reasoning paths from the fault phenomena to the candidate fault root causes in the knowledge graph. Subsequently, leveraging both the reasoning paths within the subgraphs and the fault maintenance texts, the multi-modal neural network calculates the confidence levels of the candidate fault root causes, selecting the fault root cause with the highest confidence as the reasoning conclusion. Finally, the proposed method constructs a knowledge graph based on the fault diagnosis knowledge base of key equipment for bearer network transmission provided by a domestic enterprise, and validates the efficacy of the approach using fault data employed to train the multi-modal neural network. Results demonstrate that this method effectively leverages multiple fault phenomena and their accompanying information in fault maintenance texts for accurate fault diagnosis, yielding favorable performance.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Chen et al. (2024) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: