An optimized method of module selection during software development

Abstract

The field of Software Engineering comes into the existence to avoid the damage caused by exploratory style of software development. Various steps specified during software development leads to the successful completion of the software under consideration. If the software is really not feasible then these steps will explain this in the initial investigation without wasting efforts in terms of time and person months. In this paper the focus is given on the Design and Implementation phases of the development process. A new method has been proposed to prepare a valid sequence of execution of modules to ease the work of implementation team. Convolution Neural Network has been used to find the best sequence of execution based on various coupling and cohesion parameters.