查看文章

Objectives: To estimate software development effort efficiently and easily. Methods/Statistical Analysis: Estimations at the start of the project can help identify functions involved and connect with entire the effort and implementation of a project. The proposed technique called FPBCT (Function Point Based Clustering Technique) is based on function point analysis and k-means clustering technique for achieving better predictive accuracy in software effort estimation. A live case of software development in catering is taken as a sample for study. Findings: The food service industries provide catering services for various occasions such as School functions, Colleges functions, Hospital functions, Marriage Party and many other formats, including ‘on-premises’ and ‘off-premises’ catering services. FPBCT groups the input values into three clusters based on their Euclidean distance measures of three centroids namely 3 for elementary, 5 for Medium and 8 for complex functions in a software program. In each centroid, the clusters are formed based on similarity measures ie time and lines of code and thus FPBCT effectively estimates the effort required to program software programs or modules based on function points. This research takes into account the function points in software, groups them and estimates the effort required for a project. Further, it also gives comparative efforts required in different programming languages for estimating efforts required using software function points. Application/Improvements: The proposed technique can also be adapted to object oriented programming using use cases in the future.