An experimental decision making support system for selecting appropriate online databases is presented. Three levels of metadata and primitive functions for manipulating metadata are proposed to provide basic functions of the decision making system. Primitive functions and metadata extracted from existing databases are classified into three levels from the following viewpoints: (1) degree of precision of decision making for database selection, and (2) property of metadata that gives criteria for selecting appropriate databases. From these viewpoints, the authors present a systematic design methodology for database selection in multidatabase environments. Furthermore, a primitive function set for selecting appropriate databases is exemplified to show the feasibility of the proposed system.<<ETX>>
[1]
James A. Larson,et al.
A Theory of Attribute Equivalence in Databases with Application to Schema Integration
,
1989,
IEEE Trans. Software Eng..
[2]
Gregory Piatetsky-Shapiro,et al.
An Intermediate Database Language and its Rule-Based Transformation to Different Database Languages
,
1987,
Data Knowl. Eng..
[3]
Fabio A. Schreiber,et al.
Dynamic user profiles and flexible queries in office document retrieval systems
,
1989,
Decis. Support Syst..
[4]
Gregory Piatetsky-Shapiro,et al.
An Intelligent Database Assistant
,
1986,
IEEE Expert.