IBM BPM, V8.0.1, All platforms > Authoring services in Integration Designer > Defining and transforming data > Create relationships
Relationships
A relationship correlates at least two or more semantically equivalent business objects, which are represented in different physical formats.
When source and destination business objects contain equivalent data that is represented differently, the transformation step employs a relationship. A relationship establishes an association between data from two or more business objects. Each business object is called a participant in the relationship. Each participant in a relationship is distinguished from other participants based on the function (otherwise known as role) they serve in the given relationship. Hence, a relationship will contain a list of roles. The data that you typically transform using relationships are ID numbers (such as order ID or product ID) and other values represented as codes (such as country, currency, or marital status).
For example, if a SAP application uses sequential integers for Contact IDs, and Clarify application uses generated contact codes, the SAP Contact number could be 806 and the Clarify Contact Number may be generated as A100. To transfer Contact ID data between these two applications, you can create a relationship among the SAP application's contact business object, the generic contact business object, and Clarify application's contact business object based on the ID (Key).
Relationship definitions
Relationship definitions describe what a relationship should look like. They also identify each role and specify how the roles are related. The base for a role is a data type which can be either a business object, an XSD simple type, or data types defined in interfaces.Relationships are important for mappings if no rule is applicable to cover the mapping.
For example, a product relationship of two different backend systems maps a product of System A to a product of System B. System A, however, has a different indexing system for their products than System B. Therefore, a relationship is needed as a simple mapping rule may not cover that particular mapping.
Role definitions
A role describes how entities can participate in a given relationship. That is, these role definitions are used to capture structure and constraint requirements on participating entities and their manner of participation.
For example in an employment relationship, the roles are employer and employee. A role has a set of properties/attributes specified for its semantics.
Roles have exactly one role object defined which represent the data type. Each role may have one to several key attributes defined, which represent attributes of the data type that are important for the relationship.
These role definitions are used to capture structure and constraint requirements for participating entities and their manner of participation.
For example, in a product relationship, the roles would be the product of System A and the product of System B. Each role would have defined the product number as an attribute which uniquely identifies the product in the system.
Relationship instance data
Relationship instance data is used for a static mapping of attributes of a data type. Static mappings are used for data which will not change further. This often applies to data such as zip codes, country codes, or area codes.The information in this topic deals with the pre-population of instance data, which is used in the runtime environment. The pre-population of instance data is used for lookup relationships which transform data attributes according to a static mapping. For working on relationship instance data, there is a property page in the relationship editor for the relationship in case the relationship type is set to static in the New Relationship wizard.
Relationship definition versus relationship instance
Relationship definition declares/describes what a relationship will (structure) and should (constraint system) look like. This information is encapsulated in what is known as a relationship definition. Each relationship definition has, as its constituents, role definitions.For example, a relationship definition for an employment relationship can have roles for "employer" and "employee". The relationship instance is the instantiation of these relationship definitions at run time.
For example, an instance of the relationship definition can have " John Smith" fulfill the role of employer and " Mark Johnson" fulfill the role of employee. All major relationship technologies realize the need for specifying the definition of a relationship and the need to manipulate the instances derived from the definition.
Role definition versus Role instance (Participant)
Role definition defines the entities participating in a relationship.For example, an employment relationship can have roles defined for " employer" and " employee". Role instance or participant is the instantiation of this role definition at run time.
For example, an instance of the role definition can have " John Smith" fulfill the role of employer and " Mark Johnson" fulfill the role of employee.