Selecting Software Development Life Cycle (SDLC) Models from Requirement Documents Using NLP

Abstract

The success of software system depends on many factors among which the selection of most suitable Software development life cycle (SDLC) model is the most significant. SDLC represents a framework to develop a software system through planning, analysis, design, implementation, testing, deployment, and maintenance. These activities are carried out in different series of steps and depend on the context and characteristics of the software project [1]. In this research, we will provide a view of different SDLC models with their important factors that need to be considered for their selection. Then we will propose a system to analyze the software charter document to extract useful information using NLP techniques like regular expressions. At the end, the most suitable SDLC model will be recommended for software practitioners to carry out the development process by using machine learning algorithms. We have applied 11 machine learning algorithms and achieved the highest accuracy of 90.909% with Naïve Bayes Algorithm.