Fuzzy Logic-based Expert System for Effort Estimation in Scrum Projects

This article presents an expert system based on fuzzy logic for estimating effort in scrum projects. The scrum process splits the work of the development team into short periods called sprints. A wide range of factors significantly impacts the work estimation for each task in a sprint. These factors include the development team expertise, task complexity, task size, and working environment. Typically, these factors are expressed through the use of linguistic quantifiers. As a result, this article applies fuzzy logic concepts to the development of a fuzzy logic-based expert system capable of improving effort estimation in the scrum process. A MATLAB simulation of the proposed system is used to test its behavior. The result shows that if team expertise is a beginner, task complexity is high, and if the task size is large, the story point value is complex. However, if a team is experienced, task complexity is easy, task size is small, and the work environment is best, then story point value is easy. The proposed system is tested on real data set of software projects and the results obtained are promising.