Assessing the fuzziness of general system characteristics in estimating software size

Function point analysis (FPA) is a widely used method for estimating software size. However, there are subjective elements involved in the process of function point assessment. The single estimate given by FPA does not show the confidence interval. Therefore, there is no way to assess the confidence level of the estimate. The paper introduces a fuzzified function point analysis (FFPA) model to help software size estimators to express their judgment and use fuzzy B spline membership function (BMF) to derive their assessment values. Through a case study for an inhouse software development department, the paper presents the experiences of using FFPA to estimate the software size and compares it with the conventional FPA