A Method Of Relational Database Schema Validation Based On UML Modeling

In the application of real project development,code and database structure are very tightly bound,which will change the abstraction of objects because of the advantages and disadvantages of database design.The normalization theory of relational model is the basis of logical design of relational database.The application ability of information system developers to relational normalization will directly affect the quality of the designed database system,and then affect the performance of the whole system,and reduce the cost of the system to be developed.The design of the general database structure meets the third paradigm of the relational schema,and the BC paradigm has reached the perfect level(which has completely eliminated the harmful function dependency).In a small scope,it does not affect the core business process.For the convenience of design and development,some special designs are allowed to appear.The specification is used to constrain the design.The purpose of the constraint is to eventually arrange the rationality of the design,the rationality of the final project,and the stability of the final project.This thesis is guided by the RUP software method.UML is the modeling language,combining the two,looking for the suitable view from all the UML models,recording the requirements and functions of the software process,analyzing the iterative modeling process of the system,and forming the responsibilities and responsibilities by setting up the software structure and forming the responsibility.The software architecture with clear structure is in line with the MVC mode design and realizes the idea of "high cohesion and low coupling".By transforming class diagram into database definition language(DDL),we can verify the rationality and normalization of relational schema of DDL.Taking the network bar identification system as an example,this paper verified the rationality and normalization of the third paradigm and the database relationship model,and compared the analysis results.Standardized DDL can handle data in a reasonable way to a large extent,reduce data redundancy,and ensure the correctness and stability of database design.