| Abstract |
Smart cities are the most notable examples of cyberphysical-social systems, and produce an enormous amount of industrial data related to enterprise businesses, healthcare facilities, transportation, security & surveillance, pollution and carbon emissions etc. Such multi-source data is inherently heterogeneous in nature, and therefore in order to achieve reasonably effective and efficient utilization of such data we almost always require data fusion and mining. However, such data analyses could mean the breach of the privacy of the individuals if the user data is used as it is. Privacy-preserving information consolidation and analysis approaches are therefore being developed for such use cases. In this work, an improved privacy-ensuring data fusion and service recommendation approach targeted towards users in smart city environments is presented. The approach starts with first obfuscating user-data to eliminate user-specific information (thereby ensuring user privacy), and creates equivalent time-aware indices corresponding to the user data, which are forwarded to the cloud for data fusion and service recommendation. Extensive experiments, performed on a real-world dataset, demonstrate the superiority of the proposed approach over its contemporary counterparts. |
