2021/02/11 by Mohammad Azzeh, Azzeh, Mohammad, Ali Bou Nassif +3
Computer Science · #FOS: Computer and information sciences #Software Engineering (cs.SE) #Software Engineering Research #Software Engineering Techniques and Practices #Software Reliability and Analysis Research
paper · pdf · doi:10.48550/arxiv.2102.05961
openalex publication_date 2021/02/11 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
Use Case Points (UCP) method has been around for over two decades. Although,\nthere was a substantial criticism concerning the algebraic construction and\nfactors assessment of UCP, it remains an efficient early size estimation\nmethod. Predicting software effort from UCP is still an ever-present challenge.\nThe earlier version of UCP method suggested using productivity as a cost\ndriver, where fixed or a few pre-defined productivity ratios have been widely\nagreed. While this approach was successful when no enough historical data is\navailable, it is no longer acceptable because software projects are different\nin terms of development aspects. Therefore, it is better to understand the\nrelationship between productivity and other UCP variables. This paper examines\nthe impact of data locality approaches on productivity and effort prediction\nfrom multiple UCP variables. The environmental factors are used as partitioning\nfactors to produce local homogeneous data either based on their influential\nlevels or using clustering algorithms. Different machine learning methods,\nincluding solo and ensemble methods, are used to construct productivity and\neffort prediction models based on the local data. The results demonstrate that\nthe prediction models that are created based on local data surpass models that\nuse entire data. Also, the results show that conforming the hypothetical\nassumption between productivity and environmental factors is not necessarily a\nrequirement for success of locality.\n