人工智能助推“城市交通与土地利用”课程教学 AI-Enhanced Teaching of Urban Transportation and Land Use
Authors
Zhang, Hongmou
Issue Date
2025
Educational Level
ISCED Level 6 Bachelor’s or equivalent
Curriculum Area
Geographical Setting
China
Abstract
Context: This study was conducted within an advanced undergraduate course, Urban Transportation and Land Use, at Peking University. The course examines how transport systems and land use influence each other, and how this relationship affects urban development, sustainability, and equity. Given the technical nature of the course—requiring model building, data analysis, and programming—large language models (LLMs) such as ChatGPT were introduced to help students understand complex concepts and technical methods.
Aims: The project aimed to examine how AI tools could enhance student learning in this domain, focusing on: (1) which parts of the learning process benefited from LLM support; (2) where the tools are limited; and (3) whether the use of AI creates confusion or distractions.
Methods: In the Fall 2023 semester, LLMs were integrated into two major homework assignments on transport modelling and land use analysis. Students were asked to use tools like ChatGPT or Baidu Ernie Bot, and reflected on the process, including prompt design, comparisons between AI and manual outputs, and classroom discussions. Instructors collected and analysed these reflections to evaluate the role of AI tools in learning.
Findings: Students generally found LLMs useful for understanding concepts, exploring modeling methods, and especially for coding and debugging. In the homeworks, AI tools improved efficiency in handling repetitive data tasks. However, they were less effective with “local” data or facts, and sometimes produced inaccurate or made-up results in tasks like image generation.
Implications: The study shows that AI tools can reduce barriers in technically demanding courses and support both theoretical and practical learning. It also suggests that instructors should design assignments to distinguish between tasks suitable for AI support and those requiring independent thinking.
Aims: The project aimed to examine how AI tools could enhance student learning in this domain, focusing on: (1) which parts of the learning process benefited from LLM support; (2) where the tools are limited; and (3) whether the use of AI creates confusion or distractions.
Methods: In the Fall 2023 semester, LLMs were integrated into two major homework assignments on transport modelling and land use analysis. Students were asked to use tools like ChatGPT or Baidu Ernie Bot, and reflected on the process, including prompt design, comparisons between AI and manual outputs, and classroom discussions. Instructors collected and analysed these reflections to evaluate the role of AI tools in learning.
Findings: Students generally found LLMs useful for understanding concepts, exploring modeling methods, and especially for coding and debugging. In the homeworks, AI tools improved efficiency in handling repetitive data tasks. However, they were less effective with “local” data or facts, and sometimes produced inaccurate or made-up results in tasks like image generation.
Implications: The study shows that AI tools can reduce barriers in technically demanding courses and support both theoretical and practical learning. It also suggests that instructors should design assignments to distinguish between tasks suitable for AI support and those requiring independent thinking.
Description
Keywords (free text)
higher education , artificial intelligence , LLMs , ChatGPT , urban planning , transport modelling , programming support