基于网格索引的Top-k偏好查询算法  

Top-k Preference Query Algorithms Based on Grid Index

作　　者：孙焕良[1] 姜超[1] 孙丽梅[1] 廖廷悟[2] 

机构地区：[1]沈阳建筑大学信息与控制工程学院,辽宁沈阳110168 [2]中国人民银行海口中心支行,海南海口570105

出　　处：《沈阳建筑大学学报（自然科学版）》2009年第3期579-584,共6页Journal of Shenyang Jianzhu University：Natural Science

基　　金：辽宁省自然科学基金项目(20052006);辽宁省教育厅攻关计划项目(05L354)

摘　　要：目的设计基于网格索引的Top-k偏好查询算法,提高Top-k偏好查询问题的解决效率.方法利用网格索引,采用概念划分的方法,实现基于范围查询和NN查询两种方式的Top-k偏好查询算法.结果通过真实数据集测试结果表明算法能够结合网格索引的优点,与基于R树索引的传统算法相比,在k值不断增加的情况下,查询效率提高了50%,能适应多种空间特征数据对象集合.结论网格索引可以有效处理Top-k偏好查询.To improve the problem-solving efficiency, a Top-k preference query algorithm based on grid index was designed. By indexing objects on grid structure, the conceptual partitioning method was adopted to carry out Top-k preference query algorithm which was based on the range query and the NN query. The query algorithm which was experimentally evaluated with real dataset shows that the algorithm has the advantages of the grid index, and it processes Top-k preference queries accurately and can be applied to handling multiple feature datasets in spatial database. Compared with the traditional algorithm based on R tree, the query efficiency is highly improved and the grid index is suitable for the Top-k preference query.

关 键 词：数据挖掘 Top-k偏好查询 网格索引 概念划分 

分 类 号：TP311.131[自动化与计算机技术—计算机软件与理论]