The Design And Implementation Of Quality Control System For Collaborative Crowdsourcing Testing

Crowdsourcing testing is a test model by recruiting online crowd workers to com-plete test tasks.It is usually implemented in three types:competitive,electoral,and collaborative.Collaborative crowdsourcing testing is widely used,because it recruits more crowd workers and collects more bug reports.Collaborative crowdsourcing test-ing allows crowd workers to collaborate,and bug reports are completely transparent,which can lead to malicious crowd workers.Malicious crowd workers may submit invalid bug reports,copy other's reports or review reports at will for personal bene-fit.These seriously affect the quality of bug reports and urgently need quality control strategies to ensure the quality of reports.Existing crowdsourcing quality control stud-ies are unable to evaluate the quality of reports and cannot identify malicious crowd workers,so they cannot be applied to collaborative crowdsourcing testing directly.In order to solve these problems,the thesis designs and implements a quality control sys-tem for collaborative crowdsourcing testing.The system is divided into four modules:Bug Report Validity Detection,Bug Re-port Automatic Evaluation,Feedback and Monitoring,and Bug Report Review,which guarantee the quality of bug reports from four aspects.First,the system analyzes the characteristics of bug reports,extracts the quality indicators,and detects the validity of reports according to the quality indicators,laying the foundation for the report quality evaluation and real-time feedback.Second,the system evaluates information gains of bug reports automatically,and scores the bug reports automatically with Grey Rela-tional Analysis and Entropy Weight,which implements the quality level evaluation of reports and provides a reference for managers to identify the quality of reports.Third,the system evaluates the behavior of crowd workers in real time,records the malicious behavior,and improves the quality awareness of crowd workers through malicious be-havior feedback.It also provides real-time monitoring of crowd workers*malicious behavior and collaborative relationships,helping managers identify and eliminate ma-licious crowd workers early.Fourth,the system provides an interactive and friendly manual review method.By displaying the basic attributes,information gains and au-tomatic evaluation results of the reports,the managers can review the reports quickly and fairly.In terms of technology implementation,the thesis uses the mainstream frame-works to ensure development efficiency and stability,with Angular2 as the front-end framework,Spring Boot as the server framework,and MongoDB as the database.In order to ensure system performance and scalability,Nginx is used for load balancing,and Redis is used for system cache.The quality control system passed the test and experimental evaluation.And it is running online in good condition.The system improves the quality of bug reports and the quality awareness of crowd workers,enables managers to identify the quality of bug reports and malicious crowd workers quickly,and makes it easy to review mas-sive bug reports for managers.Also,it promotes quality control study in collaborative crowdsourcing mode.