Researching On Several Key Problems About Notmalization Theory For Object-oriented Database

Normalization theory for object-oriented database was first brought forward in literature [1] by Z.Tari etc. in 1997. In contrast to conventional normalization theory for relational database, the normalization theory for object-oriented database is mainly used for database conceptual design, and independent of database logical design. In this point, it makes up for the inadequacy in logical schema, and supports the object-oriented schemata. As prospering in recently years, the theory for object-oriented database has some inadequacy and is waiting to be perfected.This paper proposed several key problems about the normalization theory about object-oriented database, and provided related solutions.The problem about candidate key of an object is an important problem in normalization theory not only for relational database but also for object-oriented database. In this paper, the problem for finding a candidate key of object was studied in details and proposed an algorithm for finding a candidate key of object.The normalization process for object-oriented database is based on the set of dependency constraints provided by the set of global dependency. So, the reduced degree of the set of global dependency affects the efficient of normalization process directly. In this paper, the problem for finding a minimal cover of a set of GD-constraints is studied. Finally, an algorithm for finding a minimal cover of a set of GD-constraints with time complexity 0(m2) is proposed.The features of the user interpretation with conflicts were studied. Once there is the user interpretation with conflicts in object, the object will never be normalized into a normalization object. In this paper, three features of the user interpretation with conflicts are extracted by studied in details.Normalization algorithm for object-oriented data model is studied. The problem of normalization for object is the kernel problem in normalization theory for object-oriented database. In this paper, the user interpretation with conflicts was studied in details. Based on this, a normalization algorithm in terms of a directed graph was provided. This normalization algorithm can identify the user interpretation with conflicts, so it improves on algorithms cited in literature [1].IIAbstractThrough the research on several key problems about normalization theory for object-oriented database , this paper makes some contributions towards the development of normalization theory for object-oriented data models.