【摘要】 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 .更多还原

【Abstract】 This paper introduces a novel approach to data mining based on Lattice Machine. This method is a way from data reduction to data mining by lattice machine. Hyper relations are a generalization of conventional database relations in the sense that sets of values are allowed as tuple entries. The collection of hyper relations can be naturally made into a complete Boolean algebra, and so for any collection of hyper tuples its unique least upper bound (lub) can be found as a reduction of it. The process is to find a set of elements i.e. interior cover in the lattice, which is closest to the least upper bound of the labeled elements, and which is such that all elements in the interior cover are “equilabeled”. This “equilabeledness” guarantees that the original labeling of the lattice is fully preserved, and is also generalized to some other originally unlabeled elements. Authors tend to find the interior cover, the subset of internal elements covered by the lub, after find the interior cover, which can in turn be used as an approach to data mining. This approach can be easily used to represent rules, and naturally avoid the missing value.更多还原