为解决目前软件行业工作量估计准确率低的问题,提出动态估计软件项目工作量的方法。首先,在项目执行前,针对工作量与规模之间的线性和非线性关系,采用基于规模的工作量估计模型对工作量进行初步估计;其次,在项目执行过程中,依据不断完善的信息对工作量估计进行调整;最后,在项目完成后,对工作量估计方法与估计结果进行评价,提出利用工作量与进度之间的幂指函数关系作为估计结果的验证指南,以提高估计结果的准确性。该方法将工作量估计作为贯穿项目始终的任务,并且将其作为一个动态的过程加以管理,为企业提高估计精度提供了一种简单、有效的方法,实现了估计方法的持续改进。
To improve the accuracy rate of software effort estimation, this paper proposed a dynamic effort estimation method. First, it adopted the size based effort estimation model to estimate the effort before project execution according to the linear and nonlinear relationship between effort and size. Secondly, it adjusted the effort based on the new informations obtained during project execution. Finally, it evaluated the estimation method and the estimation results after the project completion. In order to improve the accuracy of estimation, it took the exponential function relationship between effort and schedule as the verifica- tion guide for the proposed method. The proposed method looks effort estimation as a task that should be executed throughout the project' s whole life cycle. The proposed method provides a simple and effective method for software companies to improve their estimation accuracy, and help them achieve continuous improvement.