Research On Null Value Of Relational Database

Research on null value is an important field in relational database theory; it does not only have important theoretical values, but also has great practical values. Two-valued logic and multi-valued logic have shown much shortage in this field. As the theoretical foundation of fuzzy mathematics has set up, people introduce fuzzy mathematics to this field. Fuzzy database is a database which can handle fuzzy information, which provides a way for handling null value. At present fuzzy database technology is based on fuzzy set of L. A. Zadeh. Because vague set is more powerful than fuzzy set in the expressive power of fuzzy information, research on fuzzy database, based on vague set, has a bright future. In this paper, firstly we investigate research status of null value and fuzzy mathematics, and then the definition and property of vague set is presented. Based on vague set theory, we propose a new concept for describing null value?vague null value. A relational data model under vague null value circumstances is also proposed in this paper, and then we put forward a generalized fuzzy relational operation under vague null value circumstances.At the end of this paper, we propose a new approach to estimate null values in relational database based on rough set theory, which utilizes the data itself to find the reduced related attributes, and uses such attributes to divide the data table into different clusters, and then give the appropriate function to estimating the null value in different clusters. Because of the benefit of the rough set, the method we proposed is more objective than others.