Research On Construction Of Conversation-based Intelligent Tutoring System Based On Deep Learning

Online learning eliminates spatial barriers,enables students to learn anytime,anywhere.However,in the current online learning,the personalized learning needs of students cannot be well satisfied.Unless you connect with a teacher,otherwise,students will not be able to ask timely questions and get personalized answers.Therefore,there is an urgent need to use artificial intelligence technology to build a virtual teacher,able to help students solve cognitive barriers in time.At present,the innovative application of artificial intelligence technology in various fields is in a rapid development period.The way people interact with smart devices switch to human-computer interaction and intelligent dialogue products are loved by the public.The Conversation-Based Intelligent Tutoring System carries out teaching guidance in the way of man-machine dialogue,simulates the form of teacher-student dialogue,supports students’ independent learning,and provides learners with dialogue content that supports personalized learning.There are few researches on the Conversation-Based Intelligent Tutoring System in China.In this study,by referring to relevant literature,the research status of the Conversation-Based Intelligent Tutoring System was reviewed,and the system characteristics and construction methods of the conversational intelligent learning guidance system were expounded.The architecture of the Conversation-Based Intelligent Tutoring System is researched and designed.The first is to analyze the intention and possible entities in the learner’s discourse to realize the understanding of the student’s discourse.Secondly,the dialogue state of students is tracked and the dialogue strategies of the system are analyzed to generate teaching decisions.Third,generate system feedback in the form of a dialogue template.Fourthly,through the analysis of the relationship between entities,the knowledge graph supporting question and answer is constructed,so as to realize the architecture of the conversational intelligent learning guidance system.The first chapter of "The Doctrine of the Mean" is taken as a case to illustrate.Conversation-Based Intelligent Tutoring System is constructed by using the framework of RASA task-based dialogue system.The system build includes seven parts,The domain file、Learner discourse analysis、Process template design、Customize system actions 、 Configuration files 、 Interactive interface and Session Management.Through empirical research,the results show that the Conversation-Based Intelligent Tutoring System can produce effective teaching in the process of tutoring learners,and has a positive influence on learners’ learning.The use satisfaction survey of learners also shows that most learners recognize the quality of the content in the conversational intelligent guidance system and accept the help of the conversational intelligent guidance system to complete their learning objectives,which can have an effective impact on learning.