Requirements Elicitation for a cloud-based software with Artificial Intelligence and Machine Learning components

: a case study of the Esker Sales Order Automation Solution

Abstract

The purpose of this work is to analyze how Requirements Elicitation is led when dealing with the implementation of a software based on Machine Learning. After reviewing the existing literature about the topic, the author conducted four interviews with people who possess knowledge in Requirements Engineering. The interviews had multiple purposes: compare Requirements Elicitation techniques known in academia with the industry’s doings; find out whether dealing with a Machine Learning based software impacted Requirements Elicitation; have an insight at the overall approach to take on Requirements Elicitation i.e., methodology, stakeholders, processes, requirements reuse. The results show that there exists a gap between the focus points of academics and practitioners; respondents indicate that they do not feel like the Machine Learning components impact Requirements Elicitation; the overall process favors an Agile methodology, requirements reuse, and working with an out of the box solution complemented by customizations.

Date of Award	2021
Original language	English
Awarding Institution	University of Namur
Supervisor	Sarah Bouraga (Supervisor)