A privacy-preserving data collection model for digital community

The widespread use of mobile devices in digital community has promoted the variety of data collecting methods. However, the privacy of individuals plays an important role in data processing or data transmission, and such information should be protected. In this paper, (alpha, k)-anonymity model, a widely used privacy-preserving model, is adopted as a security frame. Then, a privacy-preserving data collection model ((alpha, k))-CM based on (alpha, k)-anonymity is proposed and the threat model is analyzed. To resist the possible attack, we propose a generalization-encryption method to achieve a desired privacy level in (alpha, k)-CM. Generalization can decrease the data size and save the resource might induce information loss in data process; while encryption can decrease information loss, however, it can cause the waste of resource. Generalization-encryption method dynamically encrypts a portion of the data with maximum information loss and adjusts the portion to balance the trade-off metric in the process of generalization. Experimental results and theoretical analysis show that this method is effective in terms of privacy levels and data quality with low resource consumption.