Research On Fusion And Evaluation Algorithms Of Curriculum Knowledge Graph For Learners

With the popularization of knowledge graph,knowledge graph has been applied to many fields,such as medical field,social field,and scientific research field.At present,many scholars have applied knowledge graph to the field of education and achieved some results,but there is still room for improvement.This paper takes the learner's curriculum knowledge graph fusion and evaluation algorithm as the research focus,maps the knowledge graph construction process to the knowledge learning process,tracks the learner's learning situation in real time,and assists the teacher to grasp the student's learning dynamics.The main research work of the thesis is as follows:(1)A method of double adjacency matrix knowledge subgraph fusion based on similarity calculation is proposed.The method first adopts the crowdsourcing method,uses the group wisdom to collect and preprocess the course knowledge,and obtains the reduced subgraphs.Secondly,the similarity calculation is performed on the labeled subgraphs,and the similarity function is obtained,and the double adjacency is combined.The individual knowledge subgraphs are merged into the group knowledge graph.Finally,the association mining method based on course community is used to obtain the potential relationship between the courses.(2)A subgraph evaluation method based on co-word analysis is proposed.The method firstly compares the individual knowledge subgraph and the fusion curriculum knowledge graph,analyzes the quality of the individual knowledge subgraph,and evaluates the knowledge subgraph.Secondly,it uses the comprehensive evaluation method to analyze the learner's learning behavior data,construct a comprehensive evaluation model of the learner,quantify the learner's learning ability,and comprehensively evaluate the learner.(3)Experimental verification and analysis of the methods proposed in this paper.Firstly,the traditional short text similarity calculation method is applied to the data set to verify the validity of the knowledge representation.Secondly,the double adjacency matrix fusion method proposed in this paper is compared with the traditionalalgorithm.Finally,the subgraph evaluation method based on co-word analysis is compared with the traditional algorithm.Experiments verify that the proposed method has good effect.