Building the Analysis Model › Performing Decomposition

Performing Decomposition

Decomposition involves breaking down activities and data into their elements. The objective of this process is to form groups of activities and data that enjoy a high level of cohesion and a low level of coupling.

• Cohesion describes how closely related the elements within each level of decomposition are. For a function, cohesion is a measure of the relatedness of its sub-functions.
• Cohesive functions generally operate on a common set of data. For a subject area, cohesion is a measure of the relatedness of its subordinate entity types.
• Cohesive subject areas are generally operated on by a common set of functions.
• Coupling describes how closely related each element in the decomposition is with elements that are not siblings.
• Loosely coupled functions have few dependencies and loosely coupled subject areas have few connecting relationships.
• A function should have more dependencies with its siblings than with functions that are not siblings.
• A subject area should have more relationships with its siblings in the decomposition than with subject areas that are not siblings.

The techniques described in "Decomposing Activities" and "Decomposing Data" tend to create highly cohesive, loosely coupled groupings of activity and data at each level of decomposition. The result is a highly stable and consistent structure that enables the efficient assignment of resources, ready coordination of development efforts, and effective project management.