Coping with uncertainties in estimating requirements development effort
Fadoua Fellir, Grace Eunjung PARK, Khalid Nafil, Lawrence Chung · 2018
Estimating requirements development effort is a critical activity in the software development process. Producing an accurate estimate, however, is extremely difficult, especially in the early stages of the software development process, where no clear and no complete description of the desired software behavior is available, hence leading to an estimate process with a high level of uncertainty. To handle this issue, in this paper, we present REEE (Requirements Elicitation Effort Estimation) - a framework for incrementally estimating software requirements elicitation efforts. We first categorize a project in terms of project's domain, different types of project parameters and requirements - functional requirements (FRs), non-functional requirements (NFRs), constraints and domain properties (DPs), while taking into consideration relationships among them. We then present an incremental approach to software requirements elicitation effort estimation that uses similar past projects in a case-based reasoning (CBR) model. In this approach, semantic similarity is quantified, once classification and normalization of requirements are done. To see the strengths and limitations of REEE, a dataset of 36(students') projects have been analyzed. Experimental results show that the use of REEE is a viable alternative to other cost estimation techniques, and can provide better results, with less uncertainty, for requirements elicitation effort estimation.