A Privacy-Preserving Trajectory Publishing Method Based on Multi-Dimensional Sub-Trajectory Similarities.

Area of Science:

• Computer Science
• Data Privacy
• Geospatial Analysis

Background:

• Location services generate vast trajectory data, necessitating robust privacy protection.
• Existing k-anonymity methods for trajectory data often rely on point similarity, creating large dummy spaces and high computational overhead.
• Current methods struggle with accurately representing trajectory similarity, impacting privacy and data utility.

Purpose of the Study:

• To propose an advanced k-anonymity trajectory privacy protection method.
• To address the limitations of existing methods by considering sub-trajectory similarity.
• To enhance the authenticity and privacy of published trajectory data.

Main Methods:

• Developed a k-anonymity method based on sub-trajectory similarity.
• Incorporated multidimensional point similarity and the area between sub-trajectories to quantify similarity.
• Utilized real historic sub-trajectories for generating indistinguishable dummy trajectories.

Main Results:

• The proposed method effectively achieves k-anonymity by generating indistinguishable dummy trajectories.
• Quantifying the area enclosed by sub-trajectories provides a more accurate spatial relationship capture.
• The approach significantly reduces the dummy trajectory space compared to point-similarity methods.

Conclusions:

• The novel sub-trajectory similarity method offers superior privacy protection for trajectory data.
• This approach enhances data quality and improves performance over existing trajectory privacy techniques.
• The method provides a more authentic and secure way to publish trajectory data using k-anonymity.