Information models define
Meta Data Services distributes a set of standard information models called the
You can use these same standard models as a framework for building new applications. You can also create new models to work with using the Meta Data Services Software Development Kit (SDK). As long as your information models are defined in UML, you can use them in Meta Data Services.
Although you can create information models programmatically, most information models are created in modeling tools like Rational Rose. Custom information models must conform to the repository API. The repository API includes abstract classes that formally describe the elements you can include in a model. If you are creating a custom model, you may want to review the repository API for more information about the type information objects that the repository engine supports.
After you define and test an information model, you can install it in a repository database. Model installation creates the storage structure for your meta data. Tools and applications that use the model can populate the storage with instance data about the model definitions.
At run time, the repository engine reads the meta data and instantiates Component Object Model (COM) objects in a tool or application that correspond to the objects, relationships, and members of your information model. The COM interfaces that you use are derived from an information model installed in a repository database. In this way, the information model is a blueprint for the COM objects that the repository engine exposes.
You can also use Extensible Markup Language (XML) to import and export meta data between platforms, tools, and applications. For more information, see Using XML Encoding.
Creating and Extending Type Information
Designing Meta Data Types Using Information Models