基于相关性分析的不完整数据函数依赖挖掘方法  

Correlation analysis for discovering functional dependencies in incomplete data

作　　者：尹诗宁 张安珍 夏秀峰 Yin Shining;Zhang Anzhen;Xia Xiufeng(School of Computer Science,Shenyang Aerospace University,Shenyang 110136,China)

机构地区：[1]沈阳航空航天大学计算机学院,沈阳110136

出　　处：《计算机应用研究》2024年第5期1368-1373,共6页Application Research of Computers

基　　金：国家自然科学基金青年基金资助项目(6210071734)。

摘　　要：函数依赖(FD)挖掘方法通常专注于发现所有满足函数依赖语法特征的结果,在数据不完整的情况下常导致大量成立但无意义的FD。针对挖掘无效FD的问题,提出基于相关性分析的不完整数据FD挖掘方法。利用概率图模型构建具有缺失值属性的概率分布,通过相关性分析捕捉属性之间的关联关系,避免枚举所有可能性,以挖掘具有统计学意义的FD。实验结果表明,该方法可以更准确地定位到有意义的FD,与最先进的FD发现方法相比,F_(1)分数平均提高1.5倍。Function dependency(FD)discovery methods typically focus on identifying all results that satisfy the syntax features of function dependencies.In the case of incomplete data,it often results in a significant number of established but meaningless FD.In response to the issue of discovering invalid FD,this paper proposed a method for incomplete data FD discovery based on correlation analysis.It constructed a probability distribution with attributes containing missing values using a probability graph model,captured the relationships between attributes through correlation analysis,avoided enumerating all possibili-ties,to discover FD with statistical significance.The experimental results show that the proposed method can more accurately pinpoint meaningful FD,more accurately,with an average F_(1)-score improvement of 1.5 times compared to state-of-the-art FD discovery methods.

关 键 词：函数依赖 相关性分析 不完整数据 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]