An Improved Probabilistic Relational Model And Its Probability To Queries

E.F.Codd proposed relational model database in the 1970's, and people began to study the method of Database and the theory of relational data, and gradually applied them to all fields. It's possible for people to handle and query massive data using database. But, There are still some questions in some fields can't be solved by current database technology. For example, There are lots of unknown and probabilistic data in the real world , Although Relational Database can store probabilistic data, but with the expansional application of traditional relationship and the development of database, these early theory which is classical relational database theory need further promote and reform.When Probabilistic Relational Model is added in the Relational Database ,because Probabilistic Relational Model can describe the uncertain data in the real world, it strengthened the ability of Relational Database expressing data,and make query results can reflect the truth of the objective world in reason more accurately. So, the research on processing method of the uncertain data under Probabilistic Relational Model has very important significance.This paper studies the Probabilistic Relational Model and its probabilistic query, the main work is as follows:Firstly, we summary the Probabilistic Relational Model based on previous studies. For the PRM model defects, we give some improvements, distinct the tuples of probabilistic relational database validly, divide it into mutually incompatible tuples and independent tuples, define the union operations on different types of tuples and solve the shortcomings of the unreasonable union effectively.Secondly, the paper describes the probabilistic query and its algorithm, gives the definition of nearest neighbor query and algorithm of R-tree-based depth-first (DF), and even improves the algorithm.Finally, a real houseagency management system is as an example, it introduces the application of Probabilistic Relational Model and its query in the database. It uses probabilitic model to establish a database model in the database system, selects the relational database management system SQL Server 2005 for the background database and uses object-oriented development tools Visual C # 2005 for system programming. Using probabilitic Relational Model in houseagency management system is a breakthrough to traditional way of housing agency services, so that housing agency can meet customer's needs better.