Clustering Objects in Object-Oriented Databases Based on Vector Space

Abstract

Clustering objects in the storage is an important problem in Object-Oriented Database Systems. Based on the relationship
sets of objects found in the system, we describe the objects as vectors in m-dimensional space, where m is the number of relationship
sets found. After that, we define a distance between pairs of objects in order to build a clustering sequence of all existing
objects in the system with their smallest total distance.

Keywords: object-oriented database, clustering objects.