Abstract

One of the recent applications of object technology is in the area of databases. One of the stumbling blocks in the commercial development and deployment of object databases is the lack of an efficient indexing technique. The properties of object databases make the task of development of an indexing technique all the more difficult. This paper discusses the development of an indexing technique for object databases. A new indexing technique based on a new structure, HC-Tree has been proposed. Performance analysis has been conducted, and experimental and analytical results indicate that the HC-Tree is an efficient indexing structure for object databases. The performance of the HC-Tree has also been compared with that of the other popular existing techniques - CH-Tree, H-Tree and hcC-Tree.

Highlights

  • During the past several years, application of objectoriented concepts has become an important topic of research in a number of disciplines in Computer Science such as databases, programming languages, and knowledge representation and computer architecture

  • Object technology has evolved in three different disciplines: first in programming languages, in artificial intelligence and in databases due to the increasing demand for sophisticated data modeling capabilities by many database applications, object databases have recently attracted a significant amount of attention

  • We propose a new structure for indexing object databases, and devise an indexing technique based on this structure

Read more

Summary

INTRODUCTION

During the past several years, application of objectoriented concepts has become an important topic of research in a number of disciplines in Computer Science such as databases, programming languages, and knowledge representation and computer architecture. Indexing techniques defined in the framework of object databases can be classified into structural - based on object attributes and behavioral - based on method execution profiles. Example: CH-Trees [K90] Two less popular techniques are nested attribute indexing - where the attribute indexed is an indirect, nested attribute of the class, and behavioral indexing - which is based on pre-computing or caching results of methods and maintaining an index. Both these methods are complex and have not been studied in detail in the literature

EXISTING TECHNIQUES
HC-Tree structure
HC-Tree indexing
Operations on the HC-tree index
PERFORMANCE ANALYSIS
CONCLUSION
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call