| Knowledge Tracing(KT)is a research hotspot in the field of artificial intelligence technology in education and an important component of smart teaching platforms.It can evaluate students’ mastery of knowledge keywords and predict future learning performance.This article proposes a knowledge tracking model based on keyword structure,and the research content mainly includes:(1)Based on the concept of results-oriented education and the theory of control science,a hierarchical model for teaching process automation is proposed by treating the teaching process as a dual closed-loop feedback control system.The inner loop of the model takes the teaching objectives of the nth teaching activity as the set values,and the evaluation reports and course portraits of students and classes as feedback signals;The outer ring takes the curriculum teaching objectives as the set value,and the evaluation report on the achievement of the curriculum objectives as the feedback signal.From the perspective of assisting teachers in teaching,it constructs an automation level for the teaching process.Similarly,the automatic driving grading is set at levels L0 to L5,providing a solution for the implementation of smart education engineering.(2)A dynamic knowledge tracking model was established based on keyword structure,and a Keywords Structure based Knowledge Tracing(KKT)model was proposed.The KKT framework uses Graph Neural Network(GNN)to map knowledge keywords into graph structures,taking into account the temporal and spatial effects of exercise sequences.It can comprehensively describe the teaching process,quantitatively evaluate learners’ keyword mastery ability,and update keyword mastery status with GNN,providing the best prediction for the next learning performance.(3)We have designed and implemented a visual intelligent teaching assistance system that can help learners establish a systematic keyword structure framework,clarify learning paths,support online learning,homework,and exams,and provide real-time evaluation reports.This system has been partially applied in real teaching processes.In summary,this article proposes a hierarchical model for teaching process automation,providing a practical path for the application of artificial intelligence technology in education and teaching;A knowledge tracking model based on keyword structure has been proposed,which can evaluate learners’ knowledge mastery ability in real-time and predict their next learning performance;At the same time,a visual intelligent teaching assistance system was designed to assist classroom teaching and applied in real teaching scenarios. |
