One of the major ways of describing the process is Data Flow Diagram. These diagrams contain, as a rule, two types of graphical objects: rectangles and arrows. Rectangles describe functions (works, processes); arrows describe data flows between these functions.
Functions are usually located from left to right in the Data Flow Diagram in the order corresponding to the sequence of their performance over time, although it is not mandatory.
If you adhere to this requirement, then the resulting diagram is a description of the process, which is similar to the description of the process using IDEF3 notation. Standard rules of decomposition are applied to the description of processes in DFD. As for the sides of the quadrangle in DFD notation, they do not have the same meaning as in IDEF0.
DFD notation is often confused with simple description of the data flow between departments. This is far not the same. Why can not simple description of flows between departments be considered as a process diagram? In every large department (for example, sales department of a large enterprise) various business processes are performed. Often these processes have different internal and external customers. That is why the chart of data flows between departments describes only data flows that cross the boundaries of the departments, but doesn’t contain all information about the external and internal changes of the data flow. That is, the Data Flow Diagram includes steps of information changes from one activity to another. In this case the description of the data flow between departments is a practically important and widely used tool.
An example of a process description in DFD can be more complicated, using the term “data storage”. This refers to any data storage, for example, a paper document, an e-fail, a database on a company’s server, etc. While modeling a process using data storages, one must remember that data can not move between the functions of the process on their own. They can only pass through certain intermediaries – data storage items, which are the same as data storage.
What are Data Flow Diagrams intended for? First of all they are needed to describe really existing data flows in the organization. Descriptions can be created by process or by function. In the first case we have business processes models in DFD format, in the second – the chart of data flow between departments. Organizational Data Flows modeling can be used to solve such problems as:
- Identification of existing data storages (text documents, fails, database management system);
- Identification and analysis of data required to perform each function of the process;
- Preparation for modeling the structure of information, so called as ERD-model (Entity-Relationship Diagrams) (IDEF1X — Integration Definition for Information Modeling)
- Selection of main and accessory organization’s business processes.
It should be noted that DFD notation can be efficiently used for describing flows of documents or resources.
Moreover, the DFD notation can be slightly improved so that data flows and resources flows can be both shown on one diagram.
In practice it is often effective to use several ways of description while processes modeling. At first, for example, we create a model in IDEF0 notation, reveal functions of the process. Then decompose the process. Reaching a certain level of specification (three-four) it becomes useful to create several charts for each detailed process: management – in IDEF0, flows of data and materials — in DFD.