Portal de Periódicos da CAPES

Particle Swarm Optimization in Small Case Bases for Software Effort Estimation

2022; Springer Science+Business Media; Linguagem: Inglês

10.1007/978-3-031-14923-8_14

1611-3349

Katharina Landeis, Gerhard Pews, Mirjam Minor,

Software Reliability and Analysis Research

To facilitate research in the field of user story effort estimation (USEE), this work analyzes the applicability of case-based reasoning to effort estimation of user stories. The analysis uses real-world user stories of a software development project of the Capgemini Group. The focus of the analysis is the development of an effort estimation tool that could generate accurate effort estimates serving the purposes of project planning and control. In addition to a classical structural CBR approach, the paper applies the weights optimization method Particle Swarm Optimization to small case bases. The highly accurate effort estimates resulting from a couple of experiments on real data show that estimation by analogy presents with the aid of an automated environment an eminently practical technique. As a consequence, this approach impacts estimation accuracy and thus success of industrial software development projects.