Study On Risk Management Of Software Development Project Based On BP Neural Network Model

With the increase of productivity and development of science and technology, modern enterprises pay more and more attention to information construction to improve efficiency, simplify processes and reduce labor costs. Corresponding software development projects emerge as the times require. Although China’s IT industry is gradually mastering the core technology,having the ability of high-tech software development, software development projects, especially the new fields not involved, rarely work out well, software development projects are always great investment and long development cycle. The larger scale, the more complex with a geometric growth, and the more uncertainty(risk). If the software is not developed successfully or fails to meet the customers requirements, huge loss will be generated to both the supply and demand. Along with the rapidly updating of information technology, enterprises demand increases and changes, which caused high risk to a certain extent to software development projects. In order to speed up the pace, many IT companies may ignore risk management, resulting in unnecessary losses. Therefore, it’s very necessary to do systematic study of software project risk management so as to find out potential risks in time, reduce the likelihood of risk, guarantee the smooth progress of the project.How to identify the potential risk before the start of a software development project, deal with the risk in the process to reduce lose, curb the expansion of the risk is the main contents of the research of software project risk management. This paper analyses and studies these questions.Firstly, this paper summarizes and analyzes the research results about software project risk management, further studies the possible risks in the development process.Secondly, we establish the risk management system of software development projects, including four parts of risk identification, risk assessment, risk response and risk monitoring. Furthermore, we do study on the most important part-risk assessment-and establish evaluation model based on BP neural network. In order to solve the problem of slow convergence speed of MATLAB simulation, a new improvement measure of adding momentum factor is put forward in this paper to improve the calculation speed and accuracy.Finally, this paper will do case analysis of YQ company’s project of network public opinion, try to identify the potential risks, seek reasonable solutions and methods, provide services for decision makers.