Neural Network Technology to Support View Integration

Abstract The most difficult and time consuming activity to perform during view integration is to find correspondences between different view specifications. Such correspondences may be the source for conflicts when integrating the views and thus must be detected and resolved. A manual inspection of the class definitions in each view and a comparison with each class definition in the other views may result in an almost endless process. To support a designer we propose a computerized tool to extract the semantics from view definitions, to transform them into a unique vector representation of each class, and to use the class vectors to train a neural network in order to determine categories of classes. The output of the tool is a ‘first guess’ which concepts in views may be overlapping or which concepts do not overlap at all. This may be of tremendous value because the designers are relieved from manual inspection of all the classes and can direct their focus on classes groupe...

• Type: Conference
• Year: 1995
• Where: ER
• Authors: Ernst Ellmer, Christian Huemer, Dieter Merkl, Günther Pernul
• Keywords: Class Definition, Database, ER 1995, Manual Inspection, Time Consuming Activity