structured systems analysis

structured systems analysis A specific technique for systems analysis that covers all activities from initial understanding of the problem through to specification and high-level design of the software system. The technique embodies four main concepts: dataflow diagrams, a data dictionary, data store structuring, and process logic representations.

The dataflow diagrams show the various processing elements in the system, and the dataflows between these processing elements and the major stores of data within the system. The processing elements are described in nonprocedural terms, typically using natural language, and a processing element from one diagram may be decomposed onto further diagrams to show greater levels of detail. A data dictionary is used to record all the various data items in the system, the constraints upon these data items, and the processing elements by which they are accessed. As the decomposition proceeds so both the data stores and the actions of the processing elements are defined in more detail. The data store structuring techniques are based upon the relational model of data and show how each data store is accessed and organized. The algorithms employed by the processing elements are defined by use of process logic representations, typically program design languages, decision tables, or “structured” natural language.

Two similar versions of structured systems analysis were developed separately by Gane and Sarson and by De Marco. The technique is intended primarily for use in traditional DP system development.