基于闭区域和开区域修剪的MTRNN过滤算法研究

Study on filter-MTRNN algorithm based on closed and open region pruning

传统的多类型反向最近邻(multiple type reverse nearest neighbor,MTRNN)查询算法没有过滤搜索空间中的数据,因此效率低下。为有效提高MTRNN查询的效率,在概述MTRNN基础上采用几何方法开发了基于R-tree的闭区域和开区域修剪方法并进一步提出了高效的过滤算法。过滤算法通过构造闭修剪区域和开修剪区域来修剪查询空间,对大数据下的MTRNN查询过滤效果十分显著。最后通过实验验证了算法的有效性。

The original MTRNN query algorithm dose not filter any data in the search space, so it is very inefficient. Based on MTRNN query, this paper developed a closed region and open region pruning strategy with R-tree, and then presented a filte- ring algorithm to reduce the search space by constructing closed regions and open regions. The filtering effect of the algorithm was obvious on large data for MTRNN queries. Finally, the experimental results demonstrate the filtering algorithm＇ s effective- ness.