Forming Software Development Team: Machine-Learning Approach

Software development team formation is a task done by skilled persons who have enough experience in mapping crew members to project tasks. Choosing software development team members according to their experience and within the amount of available budget/time is a vital task. However, the availability of a tool to suggest the best team members to the different tasks in some projects will definitely help project managers in their selections. This work is related to suggesting the most suitable skilled software development professionals to projects tasks based on some machine learning technique (Random Forest Classifier). The project manager just feeds the tool with the required tasks, and the latter suggests a ranked list of the most suitable professionals that fit each task which reflects positively on the team formation process. The experimental results conducted at the end of the work reflect the improvement of software development team formation gained comparable with the ordinary, self-experience-based one.

I. Introduction

The success of any business has several factors such as a good plan, reasonable motivation, estimating resources, and the proper selection of team members. The suitable selection of members is considered a key factor if not the most in the project success journey [1]. Software development is a leading business these days because it provides an important solution for several domains. It promotes businesses, improves sales and services, provides direct communication with customers, increases customers engagement, and helps marketing businesses [2].