Research Of Software Project Risk Management Based On Data Mining Theory

It’s a critical precondition to finish a software project smoothly by ensuring the entire software project schedule to be completed on time. One of the critical measures to guarantee smooth finish of software project is risk control in each stage of software project and some standards’development to standardize related risk control measures in the project implementation process to assure the whole project can be proceeded according to the original plan.In this paper, data mining analysis method and its application in software development projects are discussed. The risk management theory is introduced and the state-of-art of the application research about the data mining technique application in risk management system of software project is summarized. The risk management approaches in software project are described and the basic concepts and commonly used analysis approaches of data mining technology are introduced in detail. Based on current theoretical research, the extensive applications of data mining technology in real software development are analyzed. Three different specific applications of data mining algorithms in real projects are proposed on the basis of the different risks that exist in the project development process. The first one is topological constraints of timing algorithm, which is mainly used in reason review and analysis in case of errors. This algorithm can effectively improve the warning mining accuracy and efficiency of software development project. The second one is rough set analysis algorithm, which is more conducive to analyze historical data of software project risk management and find the weak links for the project to provide strategies for risk control and improve the process. The third one is decision tree model. Currently the theoretical basis of all software development projects come from software engineering theory. The analysis based on decision tree model in the beginning of the software project validates that current software risk analysis method is consistent with the software engineering theoretical basis.This thesis analyzes the common data mining analysis methods and algorithm, studies data mining applications in the actual software development projects process, designs and sets up data mining models in software project depending on the different risks. This thesis’s research is helpful for the wide data mining application in the field of communication.