M-Skyline在概率性不确定数据库上的应用  

The Application of M-Skyline over Probabilistic Uncertain Database

作　　者：曾一夫 刘泽政 周炎涛[1,2] 周旭 

机构地区：[1]湖南大学电气与信息工程学院,湖南长沙410082 [2]湖南大学信息科学与工程学院,湖南长沙410082

出　　处：《情报理论与实践》2018年第9期130-134,共5页Information Studies:Theory & Application

基　　金：国家自然科学基金项目"分布式不确定SKYLINE查询处理关键技术研究"的成果;项目编号:61472126

摘　　要：概率性的数据库由多个概率性数据元素组成,而传统的Skyline查询通常返回的是一些独立的备选推荐项,这就必然使得推荐备选项有可能是错误的。一旦这种错误发生,用户将被迫再次进行Skyline查询来寻找替代选项。这不仅导致成本损失,还损害用户体验。为了解决这个问题,文章提出了一种新的M-Skyline模型。该模型综合考虑了潜在成本损失和数据参数,并提供了有序的备选项。在真实数据库和合成数据库下对算法进行了测试。测试结果表明,所提出的算法能有效地分析处理概率性不确定数据集,并在短时间内获得M-Skyline结果。Probabilistic database is composed of multiple probabilistic data elements, and the traditional Skyline query usually returns some independent candidate recommendation items, so the recommended alternatives might be wrong. Once this error oc- curs, the user is forced to perform another Skyline query to find alternative options. This not only costs money, but also damages the user experience. In order to solve this problem, a new M-Skyline model is proposed. The proposed model comprehensively considers potential cost and data parameters, and provides ordered alternatives. The algorithm is tested in real database and synthetic data- base. The test results show that the proposed algorithm can effectively analyze and process probabilistic uncertain data sets and ob- tain M-Skyline results in a short time.

关 键 词：数据查询 数据管理 M-Skyline查询 概率数据 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]