Data Reduction Based on Hyper Relations

Data reduction makes datasets smaller but preserves classication structures of interest. It is an important area of research in data mining and databases. In this paper we present a novel approach to data reduction based on hyper relations. Our method is a generalization of the data ltering method introduced by [4] for one attribute to many attributes. Hyper relations are a generalization of conventional database relations in the sense that we allow sets of values as tuple entries. The advantage of this is that raw data and reduced data can both be represented by hyper relations. The collection of hyper relations can be naturally made into a complete Boolean algebra, and so for any collection of hyper tuples we can nd its unique least upper bound (lub) as a reduction of it. However the lub may not qualify as a reduced version of the given set of tuples. Then we turn to nd the interior cover the subset of internal elements covered by the lub. We establish the theoretical result that s...

• Type: Conference
• Year: 1998
• Where: KDD
• Authors: Hui Wang, Ivo Düntsch, David A. Bell
• Keywords: Data Mining, Data Reduction, Hyper Relations, KDD 1998, Reduction Makes Datasets