Method of differential privacy protection for web-based communities based on adding noise to the centroid of positions

The research of location privacy protection has a very important significance for the emergence and sustenance of web-based communities in the big data era. The method of differential privacy for position is a privacy model strictly for the measurement of privacy. However, under the privacy protection for multi-positions, it will bring much bigger errors by adopting the method of adding noise to protect the privacy of multi-positions. In this paper, the author puts forward a new method of differential privacy for multi-positions based on mechanism of adding noise to the centroid of positions and compares with the independent mechanism of adding noise. The experiment shows that the centroid mechanism of adding noise is better than the mechanism of independently adding noise.