Type -1 Fuzzy Logic Based System For Predicating The Best Combination Of Requirements Elicitation Techniques

Requirements Engineering (RE) is comprised of many phases. The first phase is the Requirements elicitation, which constitutes the most important phase. There have been a lot of advancements in methods that can support the elicitation phase. Each of the methods, however, has own advantages and disadvantages. These characteristics makes these methods largely varied and hence the difficulty in the selection of an appropriate method. Analysts often select methods with which they have experience or of their preference and in most cases do not consider the characteristics of the project and stakeholders. This paper proposes a type -1 fuzzy logic system which can learn the best combination of requirements elicitation techniques based on factors affecting this process and related to the characteristics of the project and stakeholders to aid novice analysts later on in the selection process. To this end, we carried out experiments on 30 projects and examined the input of 4 expert analysts who served as study participants. Our proposed system could predict a combination of methods for requirements elicitation at lower average errors. This means that the system can handle the potential uncertainties.