摘要
人工智能对高等教育影响深远,尤其在教学应用方面。以“土力学”课程为例,探讨人工智能在土木工程专业基础课教学中的应用。“土力学”课程因概念抽象、公式复杂、计算烦琐等特点,给教学和学习带来了挑战。研究选取文心一言、豆包、Kimi三种大语言模型,从交互性、知识储备、语言表达等方面分析其在课程教学中的优势与不足。结果显示,这些模型在专业术语理解、复杂计算逻辑等方面存在局限,对概念题和计算题的回答准确度有待提高。建议通过模型微调和纵向训练,结合专业课程特点优化模型,以更好地满足教学需求,为人工智能在高等教育中的应用提供参考。
concepts,complex formulas,and cumbersome calculations.This study examines three large language models—Wenxin Yiyuan,Doubao,and Kimi—analyzing their strengths and weaknesses in teaching from aspects like interactivity,knowledge base,and language expression.Results show that these models have limitations in understanding professional terminology and handling complex computational logic,with room for improvement in the accuracy of answering conceptual and calculation questions.The paper suggests optimizing the models through fine-tuning and vertical training,tailored to the course’s characteristics,to better meet teaching needs and provide references for AI applications in higher education.
作者
袁冉
程印
何毅
富海鹰
YUAN Ran;CHENG Yin;HE Yi;FU Hai-ying(School of Civil Engineering,Southwest Jiaotong University,Chengdu,Sichuan 610031,China)
出处
《教育教学论坛》
2026年第1期5-8,共4页
Education And Teaching Forum
关键词
土力学
大语言模型
课程教学
Soil Mechanics
large language model
course teaching
