Analysis Concepts And Principles


Views of Specification

Data -- entities, relationships

can be used for data modeling

Transformation (or data flow)

Structured analysis or data-flow diagrams


control flow analysis, SREM, STATEMATE

traditional analysis techniques

Object-Oriented Analysis (OOA)

emphasizes on data and transformation views

it has objects (entity and relationships) and methods (transformation)


Data View

Entity -- objects and attributes -- should we model an entity as an object or attribute?


data relationship -- one-to-one, one-to-many, many-to-many

inheritance relationship -- for modeling only

message passing relationships

Should we model relationships as objects?

relationship can be modeled as attributes


Transformation View

Data flow diagrams as functional specification

Processes and data flow among processes

Processes can be decomposed

Should we have within-object data-flow diagrams or inter-object data-flow diagrams?

How about the process of data-flow diagrams? Methods of objects?

Should we have both inter-object or within-object data-flow diagrams?


Control View

State transition diagrams, action diagrams

State diagrams can be decomposed, partitioned and abstracted.

Should we have within-object or inter-object STD or action diagrams?


Views at Global Level

If any view is missing at the global level, the specification would be difficult to understand. For example, if we have just the object model with its methods.

Difficult to figure out the control.

If we have the traditional paradigm, that is the procedural programs, it may be difficult to figure out the relationships among the data (this can be overcome by data and procedure dependence analysis).


Handling Large Problems

Partitioning, decomposition, abstraction, integration

Objects, relationships (message relationships and data relationships), scenarios (STD), methods and data-flow diagrams can support various degree of partitioning, decomposition, abstraction

For example, object partition is OK but object decomposition is unclear, object abstraction is supported by class encapsulation


Integration of Models

Integration between Object Model (OM) and Data Flow Diagram (DFD)

Each (sub) transformation must be a method of OM

Each method must be used at least once in the DFD

Integration between DFD and STD

Integration between OM and STD

Every (sub) transition within STD must be performed by methods in the OM

Each method must be used at least once in the STD

Integration among OM, DFD and STD

Integration between inter-object and within-object (STD, DFD)


Verification and Validation of Requirement Specifications

Boehm discussed four V & V criteria

Completeness -- if all of its parts are present and each part is fully developed.

Consistency -- if no conflicts exist in the specification.

Feasibility -- it is possible to create the software and the benefit exceeds the cost.

Testability -- there exists a feasible technique for determining whether the developed software satisfies the specification.


Completeness and Consistency

Internal completeness and consistency -- the specification is complete with respect to its internal structures or documents, and no conflicts exist within the specification.

External completeness and consistency -- the specification completely covers the contents in external specifications or documents, and no conflicts exist between the design specification and the requirement specification.


External Completeness and Consistency

A requirement specification is externally complete and consistent if the specification captures all user requirements of the proposed system and follows all the constraints specified by the application domain and the application itself.

A design specification is externally complete and consistent if all the requirements in the requirement specification are captured in the design specification and no conflicts exist between the design specification and the requirement specification.


Multi-Layered Framework for Testing Software Specifications

Logical separation of bugs



Reference: University of Minnesota Lecture Notes