The demand for trustworthy systems is increasingly growing as internet applications have become seamlessly integrated into our daily lives. Security trust assessment, a core component of trustworthy systems, involves the real-time collection of various user data, such as behavior patterns, access frequencies, and device information. However, the collection of extensive user information without oversight raises significant privacy concerns. To balance data availability with user privacy protection, a grouped data collection approach for trustworthy systems is proposed. This approach employs privacy-preserving data collection technology and groups users to enhance overall system efficiency. Analysis indicates that the proposed scheme can efficiently collect user data without compromising user privacy.