Lpaa：基于LangGraph的教案自动生成智能体

张豈明; 张嘉睿; 何雨晴; 张琰彬

doi:10.11959/j.issn.2096-0271

您当前的位置：

首页 >

文章列表页 >

Lpaa：基于LangGraph的教案自动生成智能体

更新时间：2026-02-27

- Lpaa：基于LangGraph的教案自动生成智能体
- Lpaa: Lesson Plan Automatic Agent Based on LangGraph
- 大数据 2025年页码：1-14
- 作者机构：
  
  华东师范大学数据科学与工程学院，上海 200062
- 作者简介：
  
  [ "张豈明（1999- ），男，华东师范大学数据科学与工程学院硕士生，主要研究方向为大语言模型应用、计算教育学。" ]
  [ "张嘉睿（2000- ），男，华东师范大学数据科学与工程学院硕士生，主要研究方向为大语言模型、计算教育学。" ]
  [ "何雨晴（2001- ），女，华东师范大学数据科学与工程学院硕士生，主要研究方向为大语言模型、计算教育学。" ]
  [ "张琰彬（1983- ），女，华东师范大学数据科学与工程学院工程师，主要研究方向为计算教育学、软件工程。" ]
- 基金信息：
- DOI：10.11959/j.issn.2096-0271
  中图分类号： TP39
- 网络首发：2025-12-17，
- 稿件说明：
移动端阅览
张豈明,张嘉睿,何雨晴等.Lpaa：基于LangGraph的教案自动生成智能体[J].大数据,

ZHANG Qiming,ZHANG Jiarui,HE Yuqing,et al.Lpaa: Lesson Plan Automatic Agent Based on LangGraph[J].BIG DATA RESEARCH,
张豈明,张嘉睿,何雨晴等.Lpaa：基于LangGraph的教案自动生成智能体[J].大数据, DOI：10.11959/j.issn.2096-0271.

ZHANG Qiming,ZHANG Jiarui,HE Yuqing,et al.Lpaa: Lesson Plan Automatic Agent Based on LangGraph[J].BIG DATA RESEARCH, DOI：10.11959/j.issn.2096-0271.

摘要

本文提出Lpaa，一种基于LangGraph构建的教案自动生成智能体（Lesson Plan Automatic Agent）。该智能体由大语言模型（Large Language Model， LLM）驱动，通过内容解析引擎、指令链生成器、前序步骤描述生成器、并行输出引擎与可控校准优化器等五个协同模块，实现知识结构提取、教学任务分解、阶段衔接生成、模块并行处理与人机协同校准。Lpaa以智能体（agent）形式运行在LangGraph的图结构上，支持任务节点化、依赖约束表达及动态调度，从而显著提升生成效率与教学内容的一致性。在自构建的多维度评估框架 pEval 下，系统平均得分 8.48/10，得到领域专家的高度认可。消融实验进一步验证了智能体各子模块对整体性能的关键贡献。研究表明，Lpaa为中文教案自动生成提供了结构化、可扩展的智能体范式，可为提升教学质量与学习体验提供有效支持。

Abstract

This paper presents Lpaa

a LangGraph-based Lesson Plan Automatic Agent designed for intelligent generation of Chinese lesson plans. Powered by a Large Language Model (LLM)

the agent coordinates five functional modules— a Content Parsing Engine

an Instruction Chain Generator

a Previous Step Description Generator

a Parallel Output Engine

and a Controllable Calibration Optimizer. These modules collaboratively support knowledge structure extraction

teaching task decomposition

stepwise contextual linkage

parallel module processing

and human-in-the-loop calibration. Operating as an agent within the graph-structured LangGraph framework

Lpaa enables task node segmentation

dependency-aware scheduling

and dynamic workflow management

thereby significantly improving generation efficiency and content consistency. Under a self-constructed multidimensional evaluation framework

pEval

the system achieved an average score of 8.48 out of 10 and received strong recognition from domain experts. Ablation studies further validate the critical contributions of each agent module to overall system performance. The results demonstrate that Lpaa offers a structured and scalable agent-based paradigm for automatic Chinese lesson plan generation

providing effective technical support for enhancing instructional quality and learning experiences.

关键词

Keywords

references

Wang Y . Research on the Evolution of Core Literacy in Chinese Language [D ] . Sichuan Normal University , 2023 . DOI: 10.27347/d.cnki.gssdu.2023.000274 http://dx.doi.org/10.27347/d.cnki.gssdu.2023.000274 .

Liang J . The Century-Long Evolution of Core Literacy in Secondary School Mathematics in China and Its Implications [D ] . Guangzhou : Guangzhou University , 2024 .

Wang T , Zhan Y , Lian J , et al . LLM-powered Multi-agent Framework for Goal-oriented Learning in Intelligent Tutoring System [EB/OL ] . arXiv: 2501.15749 , 2025 .

Liang X , Wang H , Wang Y , et al . Controllable text generation for large language models: A survey [EB/OL ] . arXiv: 2408.12599 , 2024 .

Hu T , Zhou X H . Unveiling LLM evaluation focused on metrics: Challenges and solutions [EB/OL ] . arXiv: 2404.09135 , 2024 .

Chu Z , Wang S , Xie J , et al . LLM Agents for Education: Advances and Applications [EB/OL ] . arXiv: 2503.11733 , 2025 .

Foster A , Khazanchi P , Khazanchi R . MagicSchool.ai: A Personal Assistant [C ] // Society for Information Technology & Teacher Education International Conference . AACE , 2024 : 74 – 80 .

SchoolAI [EB/OL ] . https://schoolai.com/ https://schoolai.com/ .

Curipod [EB/OL ] . https://curipod.com/ https://curipod.com/ .

Lin M P C , Chang D , Hall S , et al . Preliminary Systematic Review of Open-Source Large Language Models in Education [C ] // International Conference on Intelligent Tutoring Systems . Springer , Cham , 2024 : 68 – 77 .

姜春宇 , 韩晓璐 , 王超伦 . 智能体技术演进: 场景分析、关键技术与发展趋势 [J ] . 大数据 , 2025 . JIANG C Y,HAN X L,WANG C C L. Evolution of agent technology: scenario analysis, key technologies and development trends[J ] .BIG DATA RESEARCH, 2025.

Zeng R , Fang J , Liu S , et al . On the Structural Memory of LLM Agents [EB/OL ] . arXiv: 2412.15266 , 2024 .

Wang J , Duan Z . Agent AI with LangGraph: A Modular Framework for Enhancing Machine Translation Using Large Language Models [EB/OL ] . arXiv: 2412.03801 , 2024 .

Peng J L , Cheng S , Diau E , et al . A Survey of Useful LLM Evaluation [EB/OL ] . arXiv: 2406.00936 , 2024 .

Zhao R , Zhang W , Chia Y K , et al . Auto-Arena: Automating LLM Evaluations with Agent Peer Battles and Committee Discussions [EB/OL ] . arXiv: 2405.20267 , 2024 .

Liu Y , Iter D , Xu Y , et al . G-Eval: NLG Evaluation Using GPT-4 with Better Human Alignment [EB/OL ] . arXiv: 2303.16634 , 2023 .

Kwan W C , Zeng X , Jiang Y , et al . MT-Eval: A Multi-turn Capabilities Evaluation Benchmark for Large Language Models [EB/OL ] . arXiv: 2401.16745 , 2024 .

Shahrabi Moghadam F . Introduction to the Analytical Hierarchy Process (AHP) [J ] . Zist Sepehr Student Magazine , 2024 .

Nicaud C , Pivoteau C , Vialette S . Branch Prediction Analysis of Morris-Pratt and Knuth-Morris-Pratt Algorithms [EB/OL ] . arXiv: 2503.13694 , 2025 .

Pearce D J , Kelly P H J . A dynamic topological sort algorithm for directed acyclic graphs [J ] . Journal of Experimental Algorithmics (JEA) , 2007 , 11 : 1 .7-es.

Yu W , Zhu C , Li Z , et al . A Survey of Knowledge-Enhanced Text Generation [J ] . ACM Computing Surveys , 2022 , 54 ( 11 s): 1 – 38 .

Vaswani A , Shazeer N , Parmar N , et al . Attention is All You Need [C ] // Advances in Neural Information Processing Systems . 2017 , 30 .

Li Z , Cai J , He S , et al . Seq2seq dependency parsing [C ] // Proceedings of the 27th International Conference on Computational Linguistics . 2018 : 3203 – 3214 .

LangChain AI . Streaming [EB/OL ] . https://langchain-ai.github.io/langgraph/concepts/ streaming https://langchain-ai.github.io/langgraph/concepts/streaming , 2024-05-11 .

webup: notebooks [EB/OL ] . https://github.com/webup/notebooks/tree/main https://github.com/webup/notebooks/tree/main , 2024-05-11 .

Bradshaw J M , Feltovich P J , Johnson M . Human–agent interaction [M ] // The handbook of human-machine interaction . CRC Press , 2017 : 283 – 300 .

SmartEdu . tchMaterial [EB/OL ] . https://basic.smartedu.cn/tchMaterial https://basic.smartedu.cn/tchMaterial , 2024-05-11 .

El-Kishky A , Wei A , Saraiva A , et al . Competitive Programming with Large Reasoning Models [EB/OL ] . arXiv: 2502.06807 , 2025 .

Yang A , Yang B , Zhang B , et al . Qwen2.5 Technical Report [EB/OL ] . arXiv: 2412.15115 , 2024 .

Hu Y , Huang Q , Tao M , et al . Can Perplexity Reflect Large Language Model's Ability in Long Text Understanding? [EB/OL ] . arXiv: 2405.06105 , 2024 .

Lin C Y . ROUGE: A package for automatic evaluation of summaries [C ] // Text summarization branches out . 2004 : 74 – 81 .

Sweller J . Cognitive Load Theory [M ] // Psychology of Learning and Motivation . Vol. 55. Academic Press , 2011 : 37 – 76 .

Guizani S , Mazhar T , Shahzad T , et al . A systematic literature review to implement large language model in higher education: issues and solutions [J ] . Discover Education , 2025 , 4 ( 1 ): 1 – 25 .

Chen D , Chen R , Zhang S , et al . MLLM-as-a-judge: Assessing multimodal LLM-as-a-judge with vision-language benchmark [C ] // Forty-first International Conference on Machine Learning . 2024 .

Laskar M T R , Alqahtani S , Bari M S , et al . A systematic survey and critical review on evaluating large language models: Challenges, limitations, and recommendations [C ] // Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing . 2024 : 13785 – 13816 .

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

暂无数据