论文部分内容阅读
提出一种新的格子机数据挖掘方法 .该方法是一种从数据缩减到数据挖掘的方法 ,其中概括了传统的关系数据库的超关系被作为挖掘的对象 .超关系的集合可以被自然而然地转换为一个完整的布尔代数 .其上能够找到它的最小上确界作为缩减的结果 ,也即挖掘的结果 .该过程通过在格中寻找内部覆盖来实现数据缩减 .内部覆盖的等标注特性确保了原始数据的一致性 ,由此建立一种基于格的数据模型 .通过使用这种数据模型 ,就可以进行数据挖掘 .
A new method of gridding data mining is proposed.The method is a method from data reduction to data mining, in which the super relation of traditional relational database is summarized as the object of mining.The set of super-relationships can be converted naturally Is a complete Boolean algebra on which the smallest sums of ascertainment can be found as the result of the downscaling, that is, the result of the digging. The process achieves data reduction by finding internal coverage in the lattice. Equal labeling features such as internal coverage ensure The consistency of the original data, thus creating a lattice-based data model. By using this data model, data mining can be carried out.