Software Risk Management Model And Empirical Study

The software industry is the fastest growing industry in the world. Software project is very complex, highly technical and fast-updating; with its products'hard-to-measure abstract .All of them make the high risk of software projects. Besides, along with fast development of software industry, software projects become larger and larger, more complex and more complex, with its risk more difficult to control, leading to the failure of the software projects.Prior to completing the project, because of the lack of an effective prediction tool, enterprises should not predict the success of the project eventually; it is likely to result in major losses. Based on the summary of the software project risk management study in the past, this paper has done the work as following:Firstly,through combining literature reading in software project management ,this paper gets 50 risk factors. By prioritizing the risk factors reflected in the questionnaire surveys ,the paper obtains the Top 10 risk factors which is in line with developing characteristics of software project in our country, including poorly defined requirements, unrealistic cost and schedule estimation, insufficient plan, poor communication, process problems, staff shortage, poor change management ,poor project execution, external influence problems and implementation problems.Secondly, based on these Top 10 risk factors, according to the discriminate analysis of questionnaire data, the paper creates discriminate model which can quantitatively determine the success of a project. In the early time of the project, it can give specific information (quantitative data) and help people take measures to avoid the failure of the project finally.The data that demonstration analysis needed is obtained in the methods of questionnaire survey. There are 57 valid answers in the 60 projects. Discriminate model correctly predicted 44 of the 57 projects in the reality. the correct rate is 77.2%, which supports the hypothesis that "failure factors of unsuccessful software development projects are significantly higher than failure factors of the successful software projects."...