Research On Annotation And Representation Of Circuit Learning Resources Based On Automatic Problem Solving

In the era of "Artificial Intelligence+Education",using information technology to promote deep personalized learning has attracted much attention.The automatic and fine annotation of learning resources and personalized cognitive representation are crucial for deep personalized learning.The precision and logical annotation of learning resources directly affect the accuracy of resource recommendation,and the different representation of learning resources will also cause changes in students' cognitive psychology in the learning process.Both of the effective design of annotation and representation,can have a positive impact on students' learning.The studies on annotation and representation of learning resources mainly focus on the semantic understanding of literal labeling and the representation theory,but pay less attention to the automatic understanding or extraction of deep subject knowledge and knowledge structure behind semantics,and the automatic technology of resource presentation.Especially in physical circuits in junior high school,the automation of learning resource annotating methods and the richness of annotating results are still insufficient,and the resource representation technology also lacks automation.Therefore,it is of great significance to study the automatic fine annotation and the automatic effective representation technology of junior high school physical circuit learning resources.Based on automatic problem-solving technology,this paper proposes a framework for the automatic annotation and representation of junior high school physical circuit problem resources.The research contents mainly include the following:Research on automatic generation of knowledge labels of physics circuits problem in junior high school.Based on the syntactic semantic automatically solving technology,we get the knowledge information by understanding the meaning the problem text.The syntactic semantic model that conforming to the problem text expression of junior high school physics circuit is designed to extract the direct and implied information in the problem,so that to comprehensive understand the problem text.That is the basis of following annotation work.We design the annotation model and basic knowledge library of physical circuit.And the mapping mechanism between the knowledge information obtained from problem understanding and the basic knowledge library is established to realize the automatic generation of refined knowledge labels of the problem.Research on automatic annotation of knowledge sequence of physics circuits problem in junior high school.The knowledge sequence refers to the knowledge sequence structure formed by the logic of solving problems between the knowledge labels.In addition to the abundant knowledge labels,it should also include the logical sequence application of each discrete knowledge in the problem,that is,the correct order of knowledge labels in the process of problems solving.Compared with discrete knowledge labels,knowledge sequence contains causality and sequence between knowledge labels,which can better reflect the knowledge content,knowledge difficulty and inspection intention of the problem,and is another important annotated information of the problem.Therefore,it is necessary to annotate the knowledge sequence in the problem.In order to achieve the knowledge of sequence annotation,the correct answer process must be obtained.In this paper,we get the problem-solving process based on automatic problem solving,and extract the corresponding knowledge labels in each step of the process.According to the problem solving logic to rebuild the relationship between the knowledge labels,a group of logically related knowledge labels are recorded as a knowledge sequence.In this way,all the knowledge sequences corresponding to the problem solving process are obtained and used as the knowledge sequence label of the problem.Research on automatic representation of learning resources of physics circuits in junior high school.The different representation of learning resources in the personalized learning system is applicable to the different learning styles of students.The representation of resources completed through automation technology can reduce human resources in the style requirements of thousands of people.Therefore,under the guidance of the cognitive theory,this paper designs and implements an automatic representation framework for the learning resources of junior high school physical circuits,based on the problem understanding and solving results in the automatic problem solving.Effectively reduces the cognitive load in the learning process from the perspective of the representation of learning resources.Specifically,a color coding model is designed and implemented to highlight the information of physical elements in the text,and an image and text fusion model for completing the automatic fusion of text and image two-mode information.