基于关系模型的分布式数据库增量更新方法

Incremental Updating Method of Distributed Database Based on Relational Model

数据库更新过程中易受到噪声数据的干扰,影响分布式数据库的数据更新与查询性能,为此提出基于关系模型的分布式数据库增量更新方法。利用数学形态学与滤波器分离出分布式数据库中存在的低频信号,通过K-SVD字典处理噪声信号,完成分布式数据库数据的降噪处理。在关系模型的基础上通过边界学习算法离散化处理分布式数据库中存在的数据,并根据定性基准对数据分组,结合IUBM算法挖掘数据之间的关联规则,以此检测数据库增量。以新鲜度为依据,排序分布式数据库中存在的数据,完成分布式数据库的增量更新。实验结果表明,所提方法的去噪效果好,更新效率高,更新后数据库完整度更高。

During the update process,the database is susceptible to interference from noisy data,which affects the data update and query performance of distributed database.Therefore,an incremental updating method of distributed database based on relational model was proposed.Firstly,low-frequency signals in the distributed database were separated by mathematical morphology and filter.And then,noise signals were processed by K-SVD dictionary,so that the noise reduction of data in distributed database was completed.Based on the relational model,the boundary learning algorithm was adopted to discretize the data existing in the distributed database.Meanwhile,the data were grouped according to the qualitative criterion.Moreover,IUBM algorithm was used to mine the association rule between data,thus detecting the database increment.Based on freshness,the data in the distributed database were ranked.Finally,the incremental update was finished.The experimental results show that the proposed method has good denoising effect,high updating efficiency and higher database integrity.