AIGC-driven data preprocessing strategies for enhanced IT equipment operational insights

This study proposes an innovative AIGC-based IT equipment operating status data preprocessing framework, which solves the limitations of traditional methods in dealing with noise, missing values and inefficient feature extraction. By leveraging Generative Adversarial Networks (GANs) and Variational Autoencoders (VAEs), the framework introduces a fully automatic pipeline for adaptive noise recognition, intelligent missing data interpolation and effective feature extraction. Its key innovation lies in the integration of generative models, which eliminates the need for manual intervention and domain expertise, making it highly adaptable and user-friendly. Python-based implementation further enhances its simplicity, allowing users to implement end-to-end preprocessing with minimal configuration. Experiments on disk array datasets demonstrated their superior performance, achieving higher noise recognition accuracy (95.0% vs. 78.5%), lower missed data interpolation error (MSE: 0.12 vs. 045), and faster preprocessing (28.4 seconds vs. 72.8 seconds). The scalability and ease of use of this framework make it ideal for large-scale industrial applications, significantly improving data quality and efficiency in failure prediction. Future work will focus on extending its capabilities to multimodal data, real-time optimization and edge computing integration. This research contributes to the advancement of intelligent data preprocessing technology and provides a practical and effective solution for modern IT device management.