Abstract
Nowadays, Location-based services (LBS) are important services that take benefits from the revolution in communications. Nevertheless, user's privacy is considered as significant challenge that could impede the use of this type of services. Current Privacy-preserving techniques mainly preserve location not query privacy (i.e., query issuer identification). The untrusted LBS provider (adversary) can breach user privacy in case that it has some user background knowledge and caches queries from more than one user in the same anonymity region (group). These types of attacks use users' profiles and cached queries to predict semantically the issuer of each query. In this paper, a peer-to-peer privacy-preserving model is presented to protect the user privacy against these types of attacks taking into account the users' profiles and cached queries in the LBS server. Using this model, an inference algorithm for predicating semantically the issuer of each query and her/his underlying location is presented to check the probability that a query privacy could be breached. A set of experiments is performed to check the effectiveness of the proposed privacy-preserving model. The results show that the cached queries with semantic matching affect negatively in breaching the query and location privacy.