Article Details

Abstract

Based on the open source Large Language Model ChatGLM, a dialogue language model for teaching and research is constructed. Firstly, the basic principles of ChatGLM and its advantages in the field of Natural Language Processing are introduced. Secondly, it discusses how to use ChatGLM to construct a dialogue language model suitable for teaching and research scenarios, including key steps such as model deployment, model fine-tuning, and question answering system construction. Finally, the effect of the constructed model is evaluated, and according to the characteristics of the teaching and research industry, its application in education and scientific research scenarios is explored to achieve the goal of AI enabling teaching and scientific research. The application of Large Language Model in teaching and scientific research and student management work provides new ideas and solutions for improving the efficiency and quality of teachers' work.

Keywords

ChatGLM; teaching and research dialogue; Large Language Model; Natural Language Processing; question answering system

References

1. [1] Cui Shuang The Changes and Invariance of University Education in the AIGC Era [N]. Science and Technology Daily, August 9, 2023 (5)
2. [2] Lin Min, Wu Yuchen, Song Huan The transformation of teacher education in the era of artificial intelligence: theoretical stance, transformation methods, and potential challenges [J]. Open Education Research, 2024, 30 (4): 28-36
3. [3] Liu Yuehan, Huo Haobin, Jin Canguo Practice and exploration of building an enterprise level privatization language model assistant based on ChatGLM3 and RPA technology [J]. Architectural Design Management, 2023, 40 (12): 33-40
4. [4] Yang Xingrui, Ma Bin, Li Senyao, etc Educational text idempotent summarization method based on big language model [J]. Computer Engineering, 2024, 50 (7): 32-41
5. [5] Jin Li, Wang Dan Induction and the Bi directional Drive of Artificial Intelligence: Based on the Examination of ChatGPT [J]. Fujian Forum: Humanities and Social Sciences Edition, 2023 (7): 64-75
6. [6] Lai Ling, Wu Wei, Yang Chen, etc The Development Status, Risks, and Countermeasures of ChatGPT [J]. China Information Security, 2023 (7): 86-89
7. [7] Yin Xian, Feng Yanhong, Ye Shigen Optimization Study of Intelligent Dialogue System for Aquatic Animal Disease Diagnosis Based on ChatGLM [J]. Modern Electronic Technology, 2024, 47 (14): 177-181
8. [8] Guan Lu, He Kang, Dou Weihong Fine tuning the Large Model: Analysis of Personalized Human Computer Information Interaction Mode [J]. Press, 2023 (11): 44-51+76
9. [9] Cheng Lechao Application and Prospect of Efficient Fine tuning Techniques for Visual Large Model Parameters [J]. Artificial Intelligence, 2024 (1): 54-65
10. [10] Hu Zhiqiang, Li Pengjun, Wang Jinlong, etc Research on Policy Tool Classification Based on ChatGPT Enhanced and Supervised Comparative Learning [J]. Computer Engineering and Applications, 2024, 60 (7): 292-305

Funding

Young Innovative Talents Program for Ordinary Universities in Guangdong Province in 2023 (2023KQNCX183); 2022 Guangdong Provincial Department of Education Characteristic Innovation Project (2022KTSCX253)

About the authors

Zou Yanni (1994-), female, Han nationality, from Hengyang, Hunan, lecturer, master's degree, research direction: big data technology, information processing; He Xueqi (1995-), female, Xibe ethnic group, from Zunyi, Guizhou, lecturer, master's degree, research direction: intelligent information processing, machine learning.