A model based on information entropy to measure developer turnover risk on software project

Risk can conduce to failure of software project, and make a loss for the enterprise. It is a focus topic for software engineering researchers. Developer turnover risk is a great risk at the process of software project, yet, few scholars pay attention to it. We can use information entropy to measure effectively the uniformity of each subsystem. In this paper, we provided a quantitative measure model based on information entropy to measure developer turnover risk on software project. The more uniformly developers influence the project, the smaller the risk is. Otherwise, the core developer turnover will make a huge fluctuation to the project. We not only have analyzed the model rationality, but have provided a model case. And the data required in the model can be got in the company. The model is scientific and reasonable, which has been indicated in practice, and can be used as basis to control the turnover risk.