【摘要】 基于TPR-树(Time parameterized R-tree,TPR-树)提出的LUTPR-树采用基于主存的更新方法,更新对象的删除操作在内存中完成,故避免了TPR-树中对象更新所导致的耗时的查找操作。同时,LUTPR-树能有效利用主存资源。实验结果表明,与TPR-树相比,LUTPR-树以牺牲很小部分查询性能为代价,大大提高了更新效率,更加适用于更新密集型应用。更多还原

【Abstract】 A novel TPR-tree based indexing structure,called the(Lazy update TPR)LUTPR-tree for minimizing the updates cost,is presented.The LUTPR-tree processes object deletions in a memory-based approach,thus avoiding disk accesses for cleaning the obsolete entries during the update process.And the LUTPR-tree can efficiently use the available main-memory.Experimental results indicate that the LUTPR-tree efficiency can improve compared with the TPR-tree in update-intensive applications.更多还原