Design And Implementation Of Online Examination Platform Based On Multidimensional Cheating Detection

In 2020,the COVID-19 outbreak broke out.In order to reduce the gathering of people and reduce the risk of virus transmission,enterprises,schools and other institutions around the world began to try to transfer the work of enterprise recruitment and teaching effect evaluation to online.Internet technology has broken the constraints of time and space,brought a lot of convenience to the organization and participation of the examination,greatly improved the efficiency of recruitment and teaching effect evaluation,and reduced the cost of examination.However,compared with offline examinations,online examinations bring efficiency improvements,but also introduce fairness-related risks.At present,the widely used method of manual review of cheating behavior is inefficient,cheating behavior cannot be detected in time,and related cases emerge in an endless stream,which poses a daunting challenge to the reliability of online examination.Therefore,it is of great significance to study how to design and implement an online examination platform with effective cheating detection ability and high concurrency by using modern technology.This thesis studies the design and implementation of an online examination platform based on multidimensional cheating detection.This platform is for enterprises with recruitment needs and schools with teaching effect evaluation and examinee ability testing needs.It is used by examinees,organization administrators,organization members and super administrators.It can realize the management of users,roles,questions,examination papers,examination plans and scores and examinees’ examination participation.At the same time,this thesis proposes a multi-dimensional cheating detection scheme based on page behavior,computer vision and audio.The backend of the platform is developed based on the Spring Cloud micro-service framework.Mybatis is used as the persistence layer framework,My SQL and Redis are used to realize data storage,Kafka is used as the distributed message system,and OAuth2 is used to realize authentication.The front end of the platform is constructed with vue.js progressive framework.In terms of cheating behavior detection,this thesis constructs the anti-cheating ability of the platform through page behavior monitoring,face similarity discrimination,face occlusion detection,open-mouth speaking judgment and speech recognition.In the design and implementation of this platform,the author independently completed the requirements analysis,outline design,detailed design and implementation of the platform and system testing,and completed the front-end page construction,back-end architecture construction and business function logic functions of the online examination platform.In the construction of multi-dimensional cheating detection capability based on page behavior,computer vision and audio,this thesis completed the selection and implementation of the algorithm scheme by selecting the algorithm and verifying the inference performance of the algorithm in each feature extraction network,and embedded the algorithm into the platform to complete the overall design and implementation of the platform.This thesis fully verifies the designed and implemented online evaluation platform through functional and non-functional tests.The test results show that the platform has achieved the expected goals,can realize various management functions and test participation functions,and has effective cheating behavior detection capabilities,which provides certain ideas for the design and implementation of such platforms and the construction of cheating detection systems in the future.