This research focuses on privacy-preserving computation, federated analytics, trustworthy modeling, and intelligent decision-making for civil aviation. It considers heterogeneous and sensitive data from air traffic management, airport operations, flight trajectories, aviation communications, and equipment conditions. Cross-organizational collaborative modeling methods are developed without directly sharing raw data. Major topics include aviation data desensitization, privacy leakage assessment, access control, model security, abnormal operation detection, and explainable decision-making for air traffic flow prediction, operational risk warning, aircraft condition monitoring, and airport collaborative management.