DFD

History

Data Flow Diagrams (DFD) emerged from the structured analysis of the late 1970s. There are two main notations:

• Yourdon-DeMarco (1978) — circles for processes, open rectangles for stores
• Gane-Sarson (1979) — rounded rectangles for processes, rectangles open on one side for stores

The Gane-Sarson notation became especially common in eastern-European curricula on computer science and systems analysis.

Basic elements

• Process — a transformation of data
• Data store — where data is held between process activations
• External entity — a source or sink of data outside the system’s boundary
• Data flow — an arrow denoting the movement of data

The decomposition principle

Like IDEF0, a DFD is built hierarchically:

• Context diagram (Level 0) — the whole system and its external entities
• Top-level diagram (Level 1) — the system decomposed into its main processes
• Detail diagrams (Level 2, 3…) — each process decomposed further

Modern use

In the 2020s DFD is rarely used for business processes — BPMN and UML AD have displaced it. Yet the notation stays relevant in:

• Information-security analysis (threat modeling)
• Describing data flows during database design
• Systems-analysis courses
• Documenting legacy systems

Limitations

• Does not show temporal sequence
• Does not describe conditional logic and branching
• Weak support in modern tools