Enabling Smart Anonymity Scheme for Security Collaborative Enhancement in Location-Based Services

Abstract

Security enhancement is and always will be a prime concern for the deployment of pointof-interest (POI) recommendation services in mobile sensing environment. Recent tamper-proof technical protection such as strong encryption has undoubtedly become a major safeguard against threats to privacy in location-based services. Although the disclosure of location information could increase recommendation accuracy, the publication of trajectory data to untrusted entities could reveal sensitive details, e.g., daily routes, destinations, and favorite restaurants. In this paper, we propose a smart scheme named BUSA to approach the above problem by reconciling the tension between privacy protection and recommendation accuracy in location-based recommendation services. This scheme uses anonymizer agents positioned between the service-requesting users and location service providers; these agents operate by dividing the query information and using k-anonymity to enhance privacy protection. The scheme also utilizes clustering techniques to group users into clusters by learning their trajectory data and selects the spatial center cells as a cluster core and a benchmark for calculating recommendations via trust computing. An anonymizer coordination strategy is proposed to replace a low-performing anonymizer with one that provides stronger privacy protection for a recommendation service. The BUSA scheme adopts k-anonymity and clustering to protect privacy, and the calculated recommendation will be suitable for the cluster core that represents the entirety of users' location preferences. The security analysis reveals that the BUSA scheme can effectively protect privacy against fraudulent query requestors and the simulation results also indicate that it provides stronger privacy protection than its counterparts from the perspective of recommendation hit rate and the extent of disclosure.

An Efficient Location Privacy Protection Method for Location-Based Services based on Differential Privacy

“…Partovi et al [18] modeled a Markov decision process and introduced a new location privacy measurement method to ensure that the users' specified privacy level could be achieved in an infinite time range. The k-anonymity method mainly used in Wu et al [19] to enhance privacy protection, and used clustering technology to group users by learning their trajectory data. A graph based trajectory data representation model [20] was proposed, the similarity between trajectories was calculated using the measurement method based on edges and vertices, and similar trajectories were clustered and identified based on paths.…”

An Improved Security Framework in Health Care Using Hybrid Computing

“…A risk model and taxonomy attack are provided to elucidate the vulnerabilities in the cloud. Researchers will be able to release and immerse themselves in cloud-based Intrusion Detection approaches as a result of this work [9]. DDoS attacks have caused substantially vandalized cloud computing, also rate of failures of current DDoS attack detection approaches are comparatively large in cloud environments.…”

Semantic and Trade-Off Aware Location Privacy Protection in Road Networks Via Improved Multi-Objective Particle Swarm Optimization

“…Kuang et al [41] propose the Location Privacy Requirements, and represented it with a triplet <K, L, H>, where H denotes the privacy coefficient, the smaller it is, the narrower the scope of the candidate grid area will be, resulting in the enhancement of the utility and the decline of privacy, unfortunately, due to its tendentious implementing way, the trade-off problem is still unresolved. Combining kanonymity and clustering techniques, Wu et al [12] propose the anonymizer coordination strategy to ensures that the anonymizers always provide strong privacy protection and good service for the recommendation service. To further provide personalized privacy service, Casper's novel privacy framework in [8] is divided into two parts: the location anonymizer and the privacy-aware query processor; the former generates the cloaking regions to protect exact user location, the other is designed to deal with anonymous queries.…”

“…In this way, the anonymous process is implemented more effectively, e.g., [2] utilizes the Markov model to predict query location first before selecting anonymous cells, which reduces the interaction between users and location service provider (LSP) and improves user privacy. Besides, to provide mobile users with more effective privacy protection, many related works also focus on achieving the right trade-off between privacy and utility [7][8][9][10][11][12]. Such mechanisms generally apply the Weighted Sum Method (WSM) to balance the privacy and utility, i.e., the trade-off between privacy and utility can be achieved through parameter adjusting.…”