基于数据倾斜关联度的数据库高效挖掘方法

Database Efficient Mining Method Research Based on the Data Skew Relational

提出一种基于数据倾斜关联度的数据高效算法,首次给出数据倾斜度的概念,利用数据类间特征进行数据倾斜程度的判断,对相似数据进行类间聚类,将数据库中数据特征进行信息相似度计算,计算概念总出现次数时应累加其所有子概念的出现次数。在查询的过程中,充分考虑同一数据属性的多样性,对数据特征加入模型的数量进行约束,减少不必要的繁琐数据特征。判断符合查询条件的数据类别,根据不同的数据类别进行数据查询,实现数据结构的优化。实验结果表明,利用这种算法进行数据库查询优化,能够有效提高海量数据库信息查询的效率。

Presents an efficient algorithm based on data from the data skew incidence degree. First proposed the concept of gradient data, using the data between the characteristics of data skew degree of judgment, of similarity between classes the data clustering, the data in the database information characteristics of similarity calculation, computing concept over-all occurrences should be added when the occurrences of the concept of all children. In the process of query, give full consideration to the diversity of the same data attributes, to join model with constraints on the number of data characteris-tics, reduce unnecessary red tape data characteristics. Conform to the category of data query condition, according to the different categories of data query, optimizing the data structure. The experimental results show that using this algorithm, database query optimization, can effectively improve the efficiency of mass database information query.