Research On Software Project Cost Estimation Combined Model Of Case-Based Reasoning

For most software companies,cost estimation of software projects is a complex process.The uncertainty of the data affects the accuracy of the estimation.On the one hand,the overestimation of cost causes waste of resources and affect customer's trust in the company.On the other hand,the underestimation of cost leads to a series of problems about insufficient resources.Therefore,it is important to build a suitable software project cost estimation model and improve its accuracy and flexibility.The research object of this thesis is the cost estimation of software project.Firstly,the related concepts are introduced.Based on the introduction and analysis of the existing software project cost estimation method,a case-based reasoning technique is put forward to estimate the cost of the software project.Secondly,on the basis of the above analysis,the cost driving factors of the software project are selected,the characteristic attributes of the software project are determined,and the weight of the characteristic attributes of the case is determined by the analytic hierarchy process.Thirdly,software project cost estimation combined model based on CBR is built.Finally,we take the R software project of RF company as a specific object and make an empirical research on the CBR based software project cost estimation combined model.The attribute data in this thesis are easy to collected.In addition,the CBR based software project cost estimation combined model combines the advantages of two distance formulas,which significantly improves the accuracy and flexibility of software project cost estimation.Finally,empirical research shows that CBR based software project cost estimation combined model has significant practical value.