Scholarlink

<p>With the rapid development of information technology, network engineering education is gradually moving towards intelligence, but existing teaching content analysis and recommendation systems still have limitations in personalized evaluation. Traditional methods mostly rely on rule-driven or single modal data, resulting in incomplete construction of the Knowledge Graph (KG), low accuracy of content recommendation, and insufficient coverage of knowledge points. In response to these issues, this article combined Knowledge Reasoning (KR) and Multimodal KG with Graph Neural Network (GNN) to construct a Knowledge Reasoning and Multimodal Knowledge Graph-based Graph Neural Network (KRMKG-GNN) intelligent teaching content analysis and recommendation model based on KR and Multimodal KG. This model constructs a multimodal KG that integrates text, image, and user behavior data through GNN, and then uses KR module to dynamically evaluate students’ knowledge status and generate personalized recommendations. The experiment was validated on the learning data of 500 students, and the results showed that the model’s recommendation accuracy reached 92.7%. Compared with Graph Attention Network (GAT), the knowledge point coverage has increased by 18.5% and the error rate has decreased by 1.6. Through deep integration and reasoning analysis of multidimensional data, this model effectively enhances the intelligence level of teaching effectiveness evaluation and has broad application potential.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Intelligent analysis and recommendation of educational content in network engineering: a study on teaching effectiveness evaluation model based on knowledge reasoning and multimodal knowledge graph

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月31日

查看全文
  • Lihong Zhao

摘要

With the rapid development of information technology, network engineering education is gradually moving towards intelligence, but existing teaching content analysis and recommendation systems still have limitations in personalized evaluation. Traditional methods mostly rely on rule-driven or single modal data, resulting in incomplete construction of the Knowledge Graph (KG), low accuracy of content recommendation, and insufficient coverage of knowledge points. In response to these issues, this article combined Knowledge Reasoning (KR) and Multimodal KG with Graph Neural Network (GNN) to construct a Knowledge Reasoning and Multimodal Knowledge Graph-based Graph Neural Network (KRMKG-GNN) intelligent teaching content analysis and recommendation model based on KR and Multimodal KG. This model constructs a multimodal KG that integrates text, image, and user behavior data through GNN, and then uses KR module to dynamically evaluate students’ knowledge status and generate personalized recommendations. The experiment was validated on the learning data of 500 students, and the results showed that the model’s recommendation accuracy reached 92.7%. Compared with Graph Attention Network (GAT), the knowledge point coverage has increased by 18.5% and the error rate has decreased by 1.6. Through deep integration and reasoning analysis of multidimensional data, this model effectively enhances the intelligence level of teaching effectiveness evaluation and has broad application potential.