Levels Of Abstraction In Database Query Definition

Abstract

One of the most challenging steps in the database design and processing is the definition of queries. The planning and analysis of systems development are critical for the entire creation process. However, real computer-oriented aspects emerge in the design phase. Building up the optimum structure of an information system will determine the quality of its implementation. An important role is assigned to the database query definition, because answering queries is the final goal of an information system. Teaching different levels of abstractions and the respective strategies for query definition, such as: natural language descriptions, relational algebra expressions, QBE relational language expressions, SQL language expressions, query optimization tools, and miscellaneous recent query applications, would prepare our students for efficient and effective management of information systems.

Currently, using information systems is part of the daily activities of the global economic, social, and political environments. Many academic institutions introduce new graduate interdisciplinary programs involving a significant amount of information systems-related issues. The number of information systems users increases every day.

How do we know if our students are well prepared to develop software applications meeting the required standards? Do they have the right knowledge and skills beyond the technological level? What would be the optimum mix of theoretical and practical material for them?

Curricula designed for engineering students include courses related to the development, management, and processing of database (DB) systems and information systems (ISs). Query definition and processing is one of the most important topics considered in such courses. The entire organization of an IS is oriented towards the efficient and effective execution of corresponding queries, which extract knowledge from the data contained in the respective DB. The way of processing queries in an IS determines the success or the failure of this system2. There is a variety of textbooks3, 5, 6, 10 offering a detailed appropriate description at different levels of abstraction of query-oriented topics. Impressions obtained during professional academic forums show that college instructors are tempted to adopt textbooks, which do not include important aspects relevant to query definition and processing, issued by leading publishing houses. Usually, such books do not contain the formal description of queries based on the relational algebra (RA) and all RA-related topics. However, this is the alphabet and the heart of the relational data model used in the most popular commercial database management systems (DBMSs). Such books have their value, advantages, and readers in other cases but are not appropriate for an academic auditorium. Science is not equivalent to technology.

The main purpose of the paper is to present a vision about the right approach to teaching query definition by considering different levels of abstraction related to this problem. These levels are related to: