时态数据质量规则的研究及检测  

Research and Detection of Temporal Data Quality Rules

作　　者：黄慧[1] 李海林[2] HUANG Hui;LI Hai-lin(College of Computer Science and Engineering,Sanjiang University,Nanjing 210012,China;College of Electronic and Information Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 211100,China)

机构地区：[1]三江学院计算机科学与工程学院,南京210012 [2]南京航空航天大学电子与信息工程学院,南京211100

出　　处：《小型微型计算机系统》2021年第7期1539-1546,共8页Journal of Chinese Computer Systems

基　　金：国家自然科学基金面上项目(61671239)资助;江苏省高等学校自然科学研究项目(19KJB520049,17KJB520032)资助;三江学院校级科研基金项目(2018SJKY029)资助。

摘　　要：作为检测数据集中不一致数据的方法,函数依赖受到了广泛的关注.近年来,硬约束、等值约束、编辑规则、差分约束等被相继提出,用于发现更多的不一致数据.然而,这些约束规则仅适用于静态数据集中不一致数据的检测,而实际应用中,存在着大量随时间演化的动态数据,已有的规则忽略了具有时态语义数据的描述.该文首先提出了时态数据质量规则的形式化表达,为了提升检测效率,给出一套规则相关的性质,利用性质去除规则集中冗余规则;其次给出了不一致数据检测算法,并通过剪枝的策略对算法优化,再利用算法和不一致数据查询语言获取冲突数据;最后,通过实验验证,本文提出的方法能够检测出更多的不一致数据,经过优化后的算法执行效率较高.As a method to detect inconsistent data in datasets,functional dependencies have been widely concerned.In recent years,to better detect inconsistency, more types of rules such as hard constraints,equivalent constraints,editing rules,and differential constraints have been proposed.However,all of the proposals focus on traditional database,while ignore the semantics of temporal data which evolve with time in the real world.In this paper,we firstly propose the formalized expression of temporal data quality rules for the sake of improving the efficiency of detection,a set of properties related rules is also given,which can be used to remove redundant rules in rule sets.Secondly,corresponding algorithms are designed and optimized by pruning strategy to get conflict data with inconsistent data query language.Finally Experimental results show that the method proposed in this paper can detect much more errors,and the optimized algorithm performs effectively and efficiently.

关 键 词：时态数据质量规则 性质 检测算法 不一致数据查询语言 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]