Using Rough Sets as Tools for Knowledge Discovery

An attribute-oriented rough set method for knowledgediscovery in databases is described. Themethodis based on information generalization, whichexaminesthe data at various levels of abstraction, followedby the discovery, analysis and simplification of significant data relationships. First, an attribute-oriented concept tree ascensiontechniqueis applied to generalize the information; this step substantially reduces the overall computational cost. Thenrough set techniquesare applied to the generalized informationsystemto derive rules. Therules represent data dependenciesoccurringin the database. Wefocus on discovering hidden patterns in the database rather than statistical summaries.

• Type: Conference
• Year: 1995
• Where: KDD
• Authors: Ning Shan, Wojciech Ziarko, Howard J. Hamilton, Nick Cercone
• Keywords: Attribute-oriented Rough Set, Data Mining, KDD 1995, Significant Data Relationships, Therules Represent Data