Scholarlink

<p>With the rapid growth of spatial data, the demand for efficient and secure spatial dataset search in the cloud has become increasingly critical. However, existing search schemes operate on plaintext datasets and fail to provide privacy protection. In this paper, we propose spatial dataset range search schemes, which is the first systematic study of spatial dataset range search processing to the best of our knowledge. We design the spatial dataset-query range relevance model (SDQR) to achieve fine-grained and discriminative evaluation between query ranges and spatial datasets and propose a baseline scheme (PRDS) that supports secure search through encrypted relevance computation. To further improve efficiency, we develop an optimized scheme (PRDS+) by constructing a two-layer HNSW-GIN index and designing a dual-filtering strategy to prune irrelevant datasets. We formally analyze the security of the proposed schemes and conduct extensive experiments on three real-world spatial data repositories. The results demonstrate that our schemes achieve strong privacy guarantees while maintaining high accuracy and significantly reducing query latency and storage cost.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Privacy-preserving range-based spatial dataset top-k search processing

  • 期刊论文

  • 发表日期：

    2026年4月1日

查看全文
  • Jie Sun,
  • Hua Dai,
  • Yunhan Zhang,
  • Pengyue Li,
  • Lei Chen

摘要

With the rapid growth of spatial data, the demand for efficient and secure spatial dataset search in the cloud has become increasingly critical. However, existing search schemes operate on plaintext datasets and fail to provide privacy protection. In this paper, we propose spatial dataset range search schemes, which is the first systematic study of spatial dataset range search processing to the best of our knowledge. We design the spatial dataset-query range relevance model (SDQR) to achieve fine-grained and discriminative evaluation between query ranges and spatial datasets and propose a baseline scheme (PRDS) that supports secure search through encrypted relevance computation. To further improve efficiency, we develop an optimized scheme (PRDS+) by constructing a two-layer HNSW-GIN index and designing a dual-filtering strategy to prune irrelevant datasets. We formally analyze the security of the proposed schemes and conduct extensive experiments on three real-world spatial data repositories. The results demonstrate that our schemes achieve strong privacy guarantees while maintaining high accuracy and significantly reducing query latency and storage cost.