An Adaptive Privacy-Preserving Framework for Geospatial Data: Enhancing Accuracy and Utility in Urban Planning and Catastrophe Management

The growing need for geospatial data analysis highlight's location privacy issues. Although existing technologies such as differential privacy and location obfuscation can be relatively large, inaccurate, and difficult to handle in many applications like urban planning and disaster management, its privacy protection benefits make up for the defect. but in the research, a new privacy-preserving framework based on adaptive differential privacy model, k-anonymity and spatial cloaking is introduced. Based on geographic sensitivity, the system dynamically distributes privacy resources, thus ensuring accuracy in critical regions by choosing quality points from the data. It works to strike a balance between user privacy and system functionality by using genetic algorithm optimization and geographic analysis approaches like DBSCAN. As can be seen from the above comparisons with other systems, the experimental results show that it achieves higher levels of privacy protection than traditional methods with a lower differential privacy parameter (ε = 0.5) and larger k-anonymity k-values (k = 10). Also, the cluster quality improved phenomenally (silhouette coefficient = 0.75) and mean absolute error dropped significantly, both of which help with data utility improvement. The architecture offers a scalable, effective solution that ensures both high privacy protection and accurate representation of essential data.