This paper explores the use of the LightGBM algorithm for operator user credit assessment, a critical aspect for communication operators to establish decisions and formulate measures. The study leverages massive data provided by communication operators to develop a credit evaluation model based on the fusion of LightGBM and other machine learning algorithms. Key steps include data preprocessing, feature engineering, and the construction of multiple basic models using linear regression, decision trees, and LightGBM. The best basic models are then integrated using ensemble techniques such as Averaging, Voting, Blending, and Stacking to form a refined fusion model. The research highlights the advantages of LightGBM in handling large datasets and its superior performance in various evaluation metrics. The study also compares the effectiveness of different ensemble methods, concluding that the LightGBM-Stacking ensemble method performs the best in credit evaluation models. The findings provide a comprehensive and precise basis for operators to improve their credit assessment processes.This paper explores the use of the LightGBM algorithm for operator user credit assessment, a critical aspect for communication operators to establish decisions and formulate measures. The study leverages massive data provided by communication operators to develop a credit evaluation model based on the fusion of LightGBM and other machine learning algorithms. Key steps include data preprocessing, feature engineering, and the construction of multiple basic models using linear regression, decision trees, and LightGBM. The best basic models are then integrated using ensemble techniques such as Averaging, Voting, Blending, and Stacking to form a refined fusion model. The research highlights the advantages of LightGBM in handling large datasets and its superior performance in various evaluation metrics. The study also compares the effectiveness of different ensemble methods, concluding that the LightGBM-Stacking ensemble method performs the best in credit evaluation models. The findings provide a comprehensive and precise basis for operators to improve their credit assessment processes.