Paper
An Ameliorated Methodology for the Design of Project Data Flow Diagram
- Authors:
- Shivanand M. Handigund; Kavitha H. B
- Abstract
- The success rate of the software development projects is pathetically very low. The main causes of low success rate are attributed to i) Lack of knowledge of the project management activities; ii) Non availability of automated tools for most of the project activities; iii) Ignorance of understanding the differences between software project management and other project management activities. Thus, there is a great need to streamline the software project management activities [1]. In this paper an attempt has been made to abstract the needy components through automated methodology for the design of Project Data Flow Diagram (PJDFD). This avoids the use of arbitrary human skills in the component abstraction process. The Data Flow Diagram (DFD) components are abstracted through the design of Project work break structure (PJWBS). PJWBS is designed through the decomposition of the project work. The entire work of the project is represented as a single node and then a tree structure is generated using the designed node as the root node and then decomposing the entire work of the project randomly in three consecutive levels based on Knowledge areas (KA), Project life cycle (PLC), Software development life cycle (SDLC). The leaf nodes obtained after three tier decomposition contain the activities, both quantitatively and qualitatively. The output of these activities (the attributes defined in these activities) forms the deliverables / milestones. Some of these deliverables / milestones may have been referenced (input) to other activities. Each of these deliverables thus form the data store between the predecessor and ancestor activities. The rest of the deliverables which are not referenced in any of the activities may form intermediate or final products, results or services. Another level of datastores is identified during the pipelining process of activities. The identified activities are to be trimmed as per the semiotics of DFD [2]. During this process if the data flows in a pipelining way, then the activities are merged to form a process. Similarly if the flow of data from an activity is stacked (set of data is processed) and then flowed, we consider the stacking point as the data store and the target activities themselves as the processes. This paper discusses our proposed 22 steps methodologies. In our proposed PJDFD, we considered software requirements specification (SRS) as input and the products, results or services as output and referenced / defined attribute set as data flows.
- Keywords
- Software Requirements Specification (SRS); Enterprise Environmental Factors (EEF); Organizational Process Assets (OPA); Knowledge Areas (KA); Project Life Cycle (PLC); Software Development Life Cycle (SDLC)
- StartPage
- 175
- EndPage
- 182
- Doi