Find relationships

Your enterprise can contain a variety of data sources, physically stored in a variety of locations, and owned by a variety of people. The functions of information integration allow you to access this disparate enterprise data. The next step is to understand the data that you have access to and to determine how it all fits together. The mapping editor and the mapping model that you create help you to visualize the information that is available and how that information is related to your enterprise. The mapping model provides you with a way to use the analysis and the modeling of the relationships in other servers or database environments.

Why find relationships

With efficient design tools and metadata for the data sources that you want to integrate, you can navigate the information environment in your enterprise and understand how to use the information. Physical and integration metadata describe the physical information assets or properties of the data. This metadata is usually source-dependent, such as data type information, configuration information, or DDL syntax. This kind of information is generally used by database administrators and database developers.

Logical metadata is expressed in terms such as entity relationship diagrams or models that are defined by the Unified Modeling Language (UML). A data architect can develop abstract models about the information source or an application at a logical level in terms of entities and relationships and rules about how the structures interact.

By creating mappings with the mapping editor you can visualize the physical structures and the logical relationships that are contained in your mapping model. Without the help of the discover function, it is sometimes difficult to find subtle relationships, or relationships that are not readily apparent. When you have a large number of objects in the source and target schemas in the mapping editor, the process of finding relationships manually can be complex and time-consuming.

Discover relationships

Use the discover function to find relationships between source and target objects. These source and target objects are created from .dbm files or .xsd files. The results of the discover function appear in the center pane of the mapping editor in the form of discovered mapping objects. The Properties view shows a textual explanation of the mappings that the discover function identified.

The element mappings that are proposed with the discover function contain a number that rates the success, or confidence value, of the match. A value of 100 means that 100 per cent of the other possible matches were not as good as the match that is presented. You can specify a threshold for confidence values when you configure the discover function to determine which matches should be used.

You can annotate your mapped relationships by creating name and value pairs in the Properties view with strings. The annotations are saved with the mapping model.

The discover function provides automated relationship searches based on algorithms that you select.

Related concepts
Mapping types
Build mappings with the mapping editor
Related tasks
Setting preferences for the mapping editor
Related reference
Advanced Configuration wizard
Mapping editor

Feedback