Font Size: a A A

Design And Implementation Of Requirement Change Stratege And Bug Management In Software Development And Maintenance

Posted on:2017-03-24Degree:MasterType:Thesis
Country:ChinaCandidate:X L WangFull Text:PDF
GTID:2348330518995593Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
Through the analysis of the traditional requirements and bug management system in data records,this paper does some help in requirements similarity judge and requirements change impact measures and bug fixes priority,serious,report repeated prediction recommended so as to improve usage of the history data that recorded in traditional system and user experience,and combining predictive recommendation function with the design and implements of traditional systems.The main research points of this paper include the following four points:(1)Optimize traditional text classification algorithm for adapting to short text classification area.The traditional text classification algorithm is designed based on a large number of redundant text,but short text contains less redundant information,if we still use those traditional text classification methods,it would not work well.This paper analyzes the inherent characteristics of short text,takes actions in two aspects which are improving the text inline and key redundancy expansion,and carried out experiment which is proved that the improved algorithm can effectively improve the accuracy of short text classification.(2)Measure the impact of changes in requirements.By definition relationship between component and demand,the importance of components and influence of requirements change and similarity of demand,the paper achieves measure the impact of changes in requirements.And providing user evaluation feedback function to evaluate the system algorithm,to evaluate the system instead of artificial way,to unified evaluation criteria and reducing the amount of manual work.(3)Predicting the bug repairment in the priority,the severity and the duplication.By using data mining recommendation algorithm and opimization,this paper made a series of experiments based on the bugs from the open source system.Through prediction of the system in bug fixes,solving the new bug attribute decision affect user experience and subjective problem,and can be batch prediction,effectively improve efficiency of the bug classification prediction.(4)Based on the traditional demand and bug management system,this paper add some functions such as recommendation,prediction and recommendation of user tasks based on the user's personal label and system tags.The classification algorithms mainly from some research results in recent years,and on the basis of the the applicability to solve the problem of sparsity of the short text and effectivly improve recommendation accuracy.
Keywords/Search Tags:requirement change, software defect, text classification, bug prediction, management system
PDF Full Text Request
Related items