基于查询概率的位置隐私保护方法

现有的隐私保护技术较少考虑到查询概率、map数据、信息点（POI）语义等边信息，攻击者可以将边信息与位置数据相结合推断出用户的隐私信息，为此提出一种新的方法ARB来保护用户的位置隐私。该方法首先把空间划分为网格，根据历史查询数据计算出处于不同网格区域的用户提交查询的概率；然后结合相应单元格的查询概率来生成用户匿名区域，从而保护用户的位置隐私信息；最后采用位置信息熵作为隐私保护性能的度量指标。在真实数据集上与已有的两种方法进行对比来验证隐私保护方法的性能，结果显示该方法具体有较好的隐私保护效果和较低的时间复杂度。

Query probability-based location privacy protection approach

The existing privacy protection technologies rarely consider query probability, map data, semantic information of Point of Information (POI) and other side information, so the attacker can deduce the privacy information of the user by combining the side information with the location data. To resolve this problem, a new algorithm was proposed to protect the location privacy of users, namely ARB (Anonymouse Region Building). Firstly, the space was divided into grids, and historical statistics were utilized to obtain the probability of queries for each grid of space. Then, the anonymous region for each user was obtained based on query probability of corresponding grid to protect the user's location privacy information. Finally, the location information entropy was used as a measure of privacy protection performance, and the performance of the proposed method was verified by comparison with the existing two methods on the real data set. The experimental results show that ARB obtains better privacy protection effect and lower computation complexity.