Privacy-preserving top-k spatio-temporal keyword preference query

With the rapid advancement of location-based services and mobile devices, spatial keyword query attracts increasing attention. In this paper, we focus on a new query type known as top-$k$ spatio-temporal keyword preference query. This kind of query considers both the spatial object itself and other spatial objects in the neighborhood to return $k$ spatial objects with the highest score. These $k$ spatial objects satisfy spatial and temporal constraints, while their scores are determined by the keyword similarity of the neighboring spatial objects. We propose a scheme to enable privacy-preserving top-$k$ spatio-temporal keyword preference queries. To effectively represent temporal information, we employ time vectors to denote time periods, allowing us to assess whether the data satisfies temporal constraints based on the inner product of time vectors. Furthermore, we adopt a two-step strategy to execute the query. The first step is to find all Points of Interest (POIs) that meet the spatial and temporal constraints. The second step is to calculate the score of each POI and return the top-$k$ POIs with the highest score. To enhance query efficiency, we build a tree index structure that can achieve sub-linear query complexity. Additionally, we utilize EASPE algorithm to encrypt both the index and the query, ensuring privacy-preserving capabilities. Security analysis proves that our scheme satisfies CQA2-security. At the same time, experimental evaluation validates the query performance of our scheme.