Scholarlink

The rapid advancement of Large Language Models (LLMs) has introduced novel opportunities for enhancing project-based learning (PBL), particularly in beginner-level Python courses for students from non-computer science backgrounds. These models offer a bridge between natural language and programming logic, allowing learners to concentrate on authentic problem-solving rather than syntactic complexities. However, most current applications of LLMs in education provide fragmented, short-term support and lack coherence across the full lifecycle of project development. To address these challenges, this study proposes a Human-AI Collaborative Strategy based on the Problem-Driven Cognition and Outcome-Based Education (PDC-OBE) framework. This strategy emphasizes structured learning progression, alignment with instructional objectives, and cognitive engagement. A case study involving a sentiment analysis project was conducted to evaluate the implementation of this model. The integration of LLMs across preparation, execution, reflection, and assessment phases demonstrated significant improvements in student engagement, computational thinking, and reflective learning. The proposed framework offers a scalable, pedagogically grounded approach for integrating AI into programming education and serves as a foundational model for future development of domain-specific educational agents.

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

A Human-AI Collaborative Strategy for Project-Based Learning Using Large Language Models

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Yuan Fang,
  • Weizhen Wang,
  • Shikai Guo,
  • Mingjian Liu,
  • Xiang Li

摘要

The rapid advancement of Large Language Models (LLMs) has introduced novel opportunities for enhancing project-based learning (PBL), particularly in beginner-level Python courses for students from non-computer science backgrounds. These models offer a bridge between natural language and programming logic, allowing learners to concentrate on authentic problem-solving rather than syntactic complexities. However, most current applications of LLMs in education provide fragmented, short-term support and lack coherence across the full lifecycle of project development. To address these challenges, this study proposes a Human-AI Collaborative Strategy based on the Problem-Driven Cognition and Outcome-Based Education (PDC-OBE) framework. This strategy emphasizes structured learning progression, alignment with instructional objectives, and cognitive engagement. A case study involving a sentiment analysis project was conducted to evaluate the implementation of this model. The integration of LLMs across preparation, execution, reflection, and assessment phases demonstrated significant improvements in student engagement, computational thinking, and reflective learning. The proposed framework offers a scalable, pedagogically grounded approach for integrating AI into programming education and serves as a foundational model for future development of domain-specific educational agents.