PREDICTION OF SUITABILITY OF AGILE SOFTWARE DEVELOPMENT METHODOLOGY USING MACHINE LEARNING

Abstract

A Large numbers of software projects fail during their development phases due to high reliance of inappropriate software development methods. It is not advisable to start with a randomly chosen software development methodology for successful completion of a project within budget and target time. All the development methodologies whether belongs to agile or non-agile domain have their merits and demerits. Traditional plan-based software development methods works extremely well if the requirements are static whereas for frequently changing project requirements these methodologies are often considered as slow and insensitive. Agile methods on the other hand are considered as light-weight methods that don’t produce requirement and design documentation needs but requires intensive communication between the developers and users. Here, we present a complete framework, Development Method Selection Framework (DMSF) that provides an overall context in terms of software project parameters for exploration of project-in-hand and selection of software development method.