Research On Null-Value Relational Model In Land Information System

Null-Value relational model was used to deal with missing or uncertain information in relational database. As relational database is based on the theory of two-valued logic and classical set, when we query with this kind of imprecise information, it's possible to get wrong result. Therefore, it will lead to the inconsistency between the results and the requirement of real world. When Null-Valued model was applied into relational database, because different types of null value can describe different situations, the capacity to express data in relational database has been strengthened; this will result in more accurate reflection to the objective world. Thus, research on the theory of incomplete information database and the application of null value has grate significance.This paper aimed to do studies on the application of null value. Firstly, the definition and classification of null value was introduced. The null value and ways of handling it in SQL was also discussed in detail. Meanwhile, the influence on predicate, classifier, and aggregate function was analyzed deeply. Secondly, multi-valued logic which used to process different kinds of null value was also given. By emphasize most attentions on rules of logic operation and comparative operation in 4-valued logic, we present the algorithms of selection operation under three query statistics. Finally, by analyzing the user's requirement, the logic structure and conception structure of database was designed, and the physical model of system was given.This paper refer to three kinds of null value, it was: exist null value, non-exist null value and non-applicable null value. We handle exist null value by replace it with special value. Default value method and reasoning method were applied to process non-exist null value. As for the non-applicable null value, we can convert it to exist null value and non-exist null value according to different meaning which it express.The relational database management systems MS SQL Server2000 was choose as the support database in this land information system, high-speed prototype development method combined the programming development kit: J2EE and JSP contribute to the release of land information system . Research shows that: add null value relational model into database can describe missing information which in land database. It helps to reflect the actual situation of land using, and proved the effectiveness of null value relational model.