Creating an Estimation Model Using a COSMIC Functional Size Approximation Technique

摘要

In the actual competitive software industry, software development organizations need a better and formal estimation approach in order to increase the success rate of software projects. However, currently the estimation approach typically employed in industry is the expert judgment (`experience-based'). Using measures and estimations that continue to be based on the researchers' intuition does not contribute to obtaining successful projects nor to mature software engineering. Any organization that aims to have or develop estimation approaches needs historical data, and the majority of the organizations do not have this information. Additionally, acquisition of this information has expensive cost and time/effort consuming. This paper proposes the use of the EPCU approximation approach -that has demonstrated several benefits- aiming to create a formal database that can be employed to generate a formal estimation model with low cost and effort, improving software estimation in organizations without historical databases.