UDOOM: Uncertainty in Deductive Object-Oriented Data Modeling

Authors: Burçin BOSTAN, Adnan YAZICI

Abstract: Object-oriented data model and deductive data model are two different paradigms which are proposed to extend the traditional databases. However, both of them have some shortcomings. There are some studies which combine these two approaches and obtain a more powerful one. On the other hand, uncertainty modeling and manipulation is required in deductive object-oriented databases to handle complex objects with uncertain properties in knowledge intensive database applications, such as Expert Database Systems, Environmental Information Systems. In this study, a data model UDOOM, is proposed for representation and deduction of complex objects along with their uncertain properties. All of the null, range, and fuzzy type of uncertainty are considered. Deduction is used to be able to cope with relationships and to derive new attribute values and classes.

Keywords: Deductive object- oriented data modeling, logic and databases, uncertainty, fuzziness.