Intelligent monitoring of insult management based on machine learning and multi-source big data — Application innovation for digital economy

In the context of the digital economy, this study explores an intelligent monitoring framework for abusive supervision based on machine learning and multi-source big data. By integrating support vector machines (SVM) and random forests (RF), the research constructs a fusion model to enhance the identification and early warning of abusive behaviors within organizations. Through data collection from employee behavior records, internal feedback, and social media, the system achieves real-time analysis and high-accuracy risk prediction. Experimental results show that the integrated model outperforms single models in accuracy, precision, recall, and F1 score, especially under imbalanced data conditions. Furthermore, key variables such as job satisfaction, emotional scores, and behavioral frequency were identified as critical indicators for early detection. This study not only validates the application potential of data-driven methods in organizational behavior management but also provides enterprises with a practical intelligent tool to optimize leadership behaviors and improve employee experience.