Research On CWM-Based Relational Database Modeling Tool

The Object Management Group (OMG) is an open membership, not-for-profit consortium that produces and maintains computer industry specifications for interoperable enterprise applications. To achieve this goal, the OMG specifies open standards for every aspect of distributed computing from analysis and design, through infrastructure, to application objects and components defined on virtually every enterprise middleware platform. Model Driven Architecture (MDA) and correlative specifications, following Common Object Request Broker Architecture (CORBA) specification, aims at solving integration problems of distributed heterogeneity system. Among them, Common Warehouse metamodel(CWM) is the specification of MDA approach in warehouse and aims to drive metadata's share and interchange ability of warehouse, intellective business, and knowledge management.In this dissertation, MDA and its four core specifications: Meta Object Facility (MOF), Unified Modeling Language (UML), Common Warehouse Metamodel (CWM) and XML Metadata Interchange (XMI) are firstly introduced. Based on this, the process of three-tier architecture developed by currently MDA supporting tool is analyzed. The shortages are pointed out when the PIM model mapping to the relational PSM model in the process setting up the data centered information system. Two PIM descriptions: data PIM and Application PIM are presented to system modeling. For currently popular relation modeling tool not support CWM, so RDMT2CWM tool is developed to solve this problem. RDMT2CWM can not only save the model as CWM metadata, but also realize the function that data PIM(CWM) can be mapped to relational PSM(MS SQL Server). This software prototype is realized using Java Diagram and JDOM in Jbuilder X development environment. At last, according to the method modeling the system using two PIM presented in this dissertation, RDMT2CWM is applied in data modeling in welding management system, which makes the system easy to be expended and meta data easy to be interchanged, integrated.