基于知识图谱的跨系统电网多维数据自动挖掘

Automatic multi⁃dimensional data mining of cross system power grid based on knowledge graph

现有的基于关联分析算法、灰靶理论与云模型的挖掘技术,由于受到来源多、结构化程度复杂的数据影响,导致数据字符串挖掘效果不佳。针对上述问题,提出了基于知识图谱的跨系统电网多维数据自动挖掘方法。通过可视数据的分析与交互原理,三维融合展示跨系统电网多维数据。构建知识图谱,计算跨系统电网多维数据种类划分概率,并确定多维数据处理的目标函数。设计多维数据挖掘流程,结合SVDD算法自动分类多维数据。实验结果表明,该方法能够精准分析不同数据字符串间的依存关系,且最高召回率为98.54%,具有良好的挖掘效果。

For the existing mining technology based on association analysis algorithm,grey target theory and cloud model,due to the influence of multi⁃source and complex structured data,the effect of data string mining is poor.To solve the above problems,an automatic multi⁃dimensional data mining method for cross system power grid based on knowledge graph is proposed.Through the analysis and interaction principle of visual data,three⁃dimensional fusion shows the multi⁃dimensional data of cross system power grid.Build a knowledge map,calculate the classification probability of multi⁃dimensional data in cross system power grid,and determine the objective function of multi⁃dimensional data processing.Design multidimensional data mining process,and automatically classify multidimensional data combined with SVDD algorithm.Experimental results show that this method can accurately analyze the dependencies between different data strings,and the highest recall rate is 98.54%.