Object-Oriented Data Management

Abstract

Abstract It is increasingly recognised that good data management is essential to the petroleum industry. Relational databases are prevalent within the exploration and production industry and are increasingly applied to the storage of exploration data. In this paper we argue that the need to describe and manage complex exploration data is better served by the object oriented database management system. The OODBMS provides richer data modelling facilities which are important for representing the objects which are common in petroleum exploration, e.g. three dimensional volumes, images, etc. The facilities of the OODBMS will be introduced and contrasted with the relational database management system. An example of the use of databases in spatial data modelling will be given and the object and relational implementations contrasted. A benchmark figure for the relational database (ORACLE) and the object oriented database (ONTOS) will be given. We will show that the OODBMS is superior in speed, efficiency and data modelling capabilities. Introduction The 1990s are likely to be the decade of the object. A vast number of software vendors are using the tag "object oriented" to describe their products. Software engineering publications are continually advocating the benefits of using object orientation in the analysis, design and construction of our applications. Object oriented operating systems are starting to be released and there is considerable work being carried out on object broker technology. One more recent development (as far as commercial products are concerned) is the object oriented database management system (OODBMS). This development has come from two groups of researchers: the object oriented programming camp who have seen the need for persistent objects; and the database camp who have realised the benefits that can be gained from object orientation. The user community of OODBMSs is slowly expanding. It has been shown that the object database is very useful where the application demands a complex data model (eg CAD/CAM and knowledge engineering). They are also useful as they remove the impedance mismatch that exists between the application which uses complex data structures and relational databases that only provide tables. P. 211^