ABSTRACT
A task makes use of current data to generate fresh data, which can be either a new data item or a new value of an existing data item. For example, an XOR split (decision) node needs data to enable it to choose one of the alternative outgoing paths. A task or decision node can execute successfully only if the required data is available. The inputs data requirements of individual decision nodes and the inputs/outputs data specifications of individual tasks are assumed to be known by the process designer. However, this does not ensure that every task and decision node in a workflow can get the data needed at the time of execution and that no data anomalies can occur. Data flow analysis is a testing method for detecting improper use of data in workflow system. The data flow is an essential part of workflow execution in conjunction with control flow, and it is important to make the system work. Activities in a workflow model provide necessary data to their underlying application components and human performers correctly to identify the context of the work they are supposed to carry out, and the dataflow-analysisbased workflow design starts with identifying the set of activities needed to be included in a workflow and their input and output data through analyzing existing business processes.
