Spatial Crowdsourcing (SC) systems have emerged as an advanced application designed to match “workers” and “requesters” based on their requirements/skills and geographic locations, that named Spatial Task Matching (STM). Moreover, increasing trend of outsourcing spatio-textual data services to the cloud, directly outsourcing such data to the untrusted cloud may arise serious privacy concerns. However, existing studies rarely achieve both “efficiency” and “security” preservation simultaneously, especially in untrusted and heterogeneous environments where tasks span diverse categories and domains. To address this issue, we propose a Privacy-Preserving Weighted Bidirectional Heterogeneous Spatial Task Matching (PP-WBHSTM) framework, which treats SC as a bidirectional process, allowing both requesters and workers to upload information and perform queries equally, where previous approaches only allowed the requester to perform queries. In particular, we design an encrypted vector that enables spatial keyword queries with privacy protection, where both spatial and textual data are encrypted in a unified manner. User similarities are evaluated based on their weighted skills and locations, where each weight reflects the user’s priority level. Additionally, classification methods are employed to reduce the search complexity. This integrated approach ensures high efficiency while maintaining security and privacy, making it particularly suitable for SC systems and spatial keyword search in untrusted and heterogeneous environments. The experimental results demonstrate that the proposed PP-WBHSTM model significantly outperforms the baseline approach in all dataset sizes, achieving higher precision and recall while maintaining efficient computational performance. Moreover, it reduces the number of comparisons, confirming its superior scalability and efficiency.
Nayyeri et al. (Sat,) studied this question.