Software Requirement Prioritization using Machine Learning

Requirement engineering plays a very important role in software development life cycle (SDLC). Generally Software projects suffer with the problem of various types and categories of requirements and are also delimited by constraints like time and budget. To deal with this type of requirement complexity, project managers need to prioritize the requirements of the proposed software effectively. To decide about prioritization and consideration of a set of requirements is a strategic concern. This process is known as requirements prioritization. This paper proposes a novel requirements prioritization approach called Gradient Descent Ranking (GDRank), which combines project’s stakeholders preferences with Functional and Non Functional requirements, their ordering and approximations are estimated through machine learning techniques. For validation purpose the proposal is compared with various other prominent requirement prioritization methods. Keywords-Quality Function Deployment (QFD), Requirements Prioritization, Machine Learning