| For decades software managers have been using formal methodologies such as the Constructive Cost Model and Function Points to estimate the effort of software projects during the early stages of project development. While some research shows these methodologies to be effective, many software managers feel that they are overly complicated to use and essentially trades one estimation model for another. Instead, a larger majority of software managers have turned to a less formal methodology called "expert estimation" or "estimation by analogy". These methodologies have shown to be extremely accurate when used by a manager who has a significant amount of domain experience. While expert estimations have proven to be quite effective their accuracy relies on vague, imprecise, or incomplete information available at the start of a software project. In addition, they rely on the tacit knowledge held by the manager, which makes duplicating their results, by other managers difficult to do. This research proposes a methodology where expert estimation in conjunction with fuzzy logic is used to determine the effort required to develop a real world software engineering project. |
