Exploring and validating the contributions of real-world knowledge to the diagnostic performance of automated database design tools

Abstract

Automated database design tools employ knowledge-based systems technology in order to provide intelligent support to humans during the process of database analysis and design. However, the level to which these tools can simulate the diagnostic capabilities of human designers when performing a design task remains in question. Human designers employ what might be called "knowledge of the real world" in carrying their design activities; such knowledge is employed by only a few automated database design tools. Therefore, in recent years, there have been a number of attempts to develop tools that are capable of exploiting such real-world knowledge. It has been claimed that the use of such knowledge has the potential to increase the diagnostic performance of automated database design tools. However, to date, little if any formal exploration and validation of this claim has taken place. This paper presents our activities in exploring and validating the implications for exploiting three approaches facilitating the use and exploitation of real-world knowledge in the diagnostic performance of database design tools. The results obtained have demonstrated that the improvement of certain aspects of diagnostic performance has been achieved. However, the extent to which these aspects have been attained and subsequently statistically validated varies.