Extracting Domain Semantics for Knowledge Discovery in Relational Databases

Our research investigates howdomainsemantics are discovered froma relational database at a high level of automation. Thediscoveries are then represented as the extendedEntityRelationship schema[2] (i.e., conceptual schema)and integrity constraints of the database, whichcan support and drive the subsequentknowledgediscovery in this database. This paper presents a discovery process whichobtains an extended Entity-Relationship schemafrom a relational database by analyzing not only data instances but also the executable data schemaof'the database. In addition, our research appears to be the first ones to address the problemof erroneousdata for discovering inclusion dependencies.Aninteractive knowledge-basedsystem, the KnowledgeExtraction System(KES), has been developed to perform the discovery process. KESdemonstrates howknowledge-basedsystems technology can be applied to supportthe workof knowledgediscoveryin a database. It also illustrates that the discovery process can be implemen...

• Type: Conference
• Year: 1994
• Where: KDD
• Authors: Roger H. L. Chiang, Terence M. Barron, Veda C. Storey
• Keywords: Data Mining, Discovery Process, KDD 1994, Relational Database, Research Investigates Howdomainsemantics