Access methods on aggregation of object-oriented database

Abstract

In object-oriented databases, an object belongs to a class which consists of a set of attributes and the values of these attributes are allowed to contain a set of objects that belong to other classes; thus, forming an aggregation hierarchy of classes. To expedite the evaluation of a query on a nested class, numerous researches have been undertaken. However, the majority of the methods proposed are based on tree indexing structures. In this paper, we present three organizations based on multiattribute hashing for use in the efficient evaluation of a query in an object-oriented database. Furthermore, we consider the possibility of employing signature files to speed up retrieval for these access methods. We develop detailed models for a representative organization from each of the three groups i.e. tree-based, hash-based and signature files. Using the models, we evaluate the storage cost and retrieval cost of these groups of organizations.