The Dependency Matrix is a quick method for representing dependencies compactly. Diagrams, UML elements, and extended UML elements serve as row and column entries. The cells in the matrix show where these elements are associated, that is related.
Dependency matrices can show dependency criteria, such as
BPML, SysML, UPDM, and other DSL relationships
Semantic dependencies (dependencies through properties)
Relationships through tags (tags allow for relating UML elements that cannot be represented on the same diagram)
Transitive relationships (through the OCL or meta chain)
Dependency matrices are useful for
Quickly visualizing and creating dependency criteria.
Compactly visualizing relations of a big system. Such system relations cannot be represented by a diagram on a single sheet of paper, as the diagram is very big.
Studying relations from a particular scope and type of element by filtering the unimportant data. Hierarchical columns allows for browsing directly in the created matrix through the model scope and visualize the required area.
Showing relations that cannot be represented on diagrams, such as a representation (a class by the lifeline), a behavior representation in other diagrams, an operation representation by Call Behavior Action, Use Case relations with describing activities through the property Owned Behavior, etc. The semantic dependency matrix is needed for the deeper model analysis. The Dependency Matrix allows the representation of any kind of relations through element properties.