The Research And Implementation Of Diagnostic Evaluation Model And Test Paper Generation Algorithm In College English Diagnostic Practice System

With the widespread popularity of the Internet,online learning had become an important learning way for everyone and a hot research field for many universities and companies at home and abroad.However,after trying some online exercise platforms,the author found that these platforms still repeated the traditional mode of doing exercises and checking answers,which did not provide personalized diagnostic evaluations and targeted exercises recommendations.At the same time,the current College English Diagnostic Test System also had these problems.Therefore,this dissertation researched the diagnostic evaluation by data mining with large amounts of data collected from database,and put forward two test paper generating algorithms based on the diagnostic evaluation and implemented to perfect the diagnostic evaluation functions of College English Diagnostic Test System.First,this dissertation researched and designed the diagnostic evaluation model.Then,the author analyzed a large amount of data collected from database in three aspects:using Student-Problem Chart Analysis method to construct a learning condition evaluation model,using data mining to build a question types association model and using machine learning to construct the prediction of CET-4 grades model.The diagnostic evaluation model was finally obtained by merging these three sub models and the existing knowledge points association rules.Second,the diagnostic evaluation model was introduced,this dissertation designed a test paper generating algorithm based on learners' learning condition and an another test paper generating algorithm based on the question types association rules,and we carried on the related validation and analysis in three aspects,including composing time,exercises recommendations and grade improving.Finally,based on the requirement analysis,this dissertation used diagnostic evaluation model and new test paper generating algorithms to design and implement the diagnostic evaluation subsystem in the College English Diagnostic Test System.The diagnostic evaluation model proposed in this dissertation can accurately evaluate learning condition and diagnose learning obstacles dynamically.The comparison experiment showed that two test paper generating algorithms based on the diagnostic evaluation model can give better learning guidance and exercises recommendations,and it can also effectively improve students' grades,which had extensive application prospect and research significance.