Dispersed dummy selection approach for location-based services to preempt user-profiling

摘要

Location-based services (LBS) provide specific personalized services based on the location information provided by the user. Disclosing the users' private data is a challenging issue prevailing in LBS. Users become victims of online user-profile disclosure risks when they continuously utilize the services and attempt to disclose private data. The proposed dispersed dummy selection-based approaches such as dispersed dummy selection including actual location (DDSIA) and dispersed dummy selection excluding actual location (DDSEA) avert user-profiling issue. The dummy locations are chosen based on the auxiliary information of the location, user, and physical dispersion. The auxiliary information about the location is fetched from the local fog servers, and auxiliary information about the user is stored and retrieved from the cloud storage. An anonymous circular area is considered to ensure the dummy locations are dispersed, and the user-location correlations get dissipated. The detailed analysis of the proposed approaches is explored with the real-world map data, and its effectiveness is verified with simulations using Matlab. In addition, the qualitative analysis tests the resistance of the dummies against human observations. The results obtained substantiate that the proposed approaches are efficient in terms of dispersion degree and resistance to human observations that avert user profiling.