Computer systems process data. One way of representing a system is to use a data flow diagram.

This type of diagram shows how the data flows through the system, and what data stores are used. It does not define what type of data storage is used, or how the data is stored. this type of detail can be determined at a later stage.

Symbols used :

External Entity - data source or data destination - ie where the data comes from or where it goes to...eg timesheet, customer, accountant,etc...

 
Process - an operation performed on the data.

Data store - this could, for example, represent a file held on disc or magnetic tape, a batch of input documents or a report.

 Data flow - the arrow represents movement of data between entities, processes or data stores. The arrow should be labelled to describe what data is involved.

When drawing data flow diagrams, you should stick to the following conventions:

• do not draw data flow lines directly between data stores and external entities. There should be a process box between them to show the operation performed.
• most processes will have input data and output data...so state what they are.
• label all the data flow lines so that it is clear what data is being transferred.

Example of a Data Flow Diagram

The payroll system in a certain company may be described as follows:

At the end of each week an employee hands in a time sheet that is sent to the computer centre showing how many hours the employee has worked. The  employee's data (pay rates, NI Number etc) is stored on the Employees database and is used to process the weekly payroll, which produces a payslip that is given to the employee. The details of the payment are stored in the Accounts database.

A simple data flow diagram might be as follows: