Integrating AI Techniques in Requirements Analysis

Abstract

Requirements Analysis (RA) remains one of the most central processes in requirements engineering. It is a process of evaluating and discovering possible structures to create a contracted set of broad and reliable requirements. The major aim of requirements analysis is to produce a requirements specification document with great quality. The analysis of study exposes that experts have arranged major supports by developing various methods/tools/framework/techniques of requirement analysis process. Though, one of the major problems faced by developers is poor communication and frequently changes in requirements. These issues may lead to incompetent outcome and termination of the system development. The previous investigation exposes that Artificial Intelligence (AI) methods may support in this by restricting alterations in requirements and to propose effective communication between designers and users. The purpose of this work is to categorize the challenges in every stage of the requirements analysis and incorporation of AI techniques to solve these known challenges. Moreover, the research also determines the association between such challenges and their potential AI answer/s through Venn-Diagram. Prior studies expose that more than one AI technique available for some of the challenges, and some of the challenges are still open for further research, no AI techniques has been reported yet. Keeping in observation the significance of the area, foremost analysis methods and their related issues have already been recognized in one of our prior papers. This study is an addition of our prior effort and here, an attempt is made to incorporate and describe AI techniques in various requirements analysis techniques.