1 Introduction

It highlights the importance of International Chinese Language Education in disseminating Chinese culture, points out the existing problems in the digital resource development of International Chinese Language Education, such as resource homogeneity and low level of intelligence, and proposes the necessity of utilizing generative artificial intelligence to promote the digital transformation of International Chinese Language Education. It also introduces the interdisciplinary perspective of this research and the exploratory direction based on the community theory model.

2 Main Research on AI-Enabled International Chinese Language Education

It primarily explores the application research of artificial intelligence in International Chinese Language Education. Firstly, it analyzes the research and practices of the digital transformation of International Chinese Language Education, including the development of digital resources, teaching method research, and interdisciplinary collaborative innovation. Then, it elaborates on the application and breakthroughs of artificial intelligence in second language learning, such as providing personalized learning content and real-time feedback through natural language processing and machine learning technologies to enhance learners' language abilities. Finally, it discusses the potential and value of generative artificial intelligence in Chinese writing instruction, such as creating text content based on large language model technology to provide personalized writing feedback and guidance, thereby promoting learners' writing skills. It also points out the current challenges, including the shortage of digital resources and the development of advanced Chinese writing large language models.

3 Research Framework for Generative AI-Enabled Digital Writing Platforms in International Chinese Language Education

Firstly, it elaborates on the background of international friends' needs for learning Chinese writing skills and the opportunities for applying generative artificial intelligence in Chinese writing instruction. Then, it introduces the practical exploration of building a specialized Chinese writing agent based on the CoI theory model, including constructing a private knowledge base and training a vertical large model. Finally, it presents future prospects for the application of agent teaching assistants, continuous iterative optimization, and promotion and application, aiming to enhance the writing abilities of Chinese learners and promote the collaborative development of the teaching community.

4 Construction and Practical Exploration of a Specialized Chinese Writing Agent

This chapter primarily explores the construction process, capability testing, and comparisons with general large models and other agents of the specialized Chinese writing agent. In terms of the construction process, ChatGLM was selected as the base model, with basic information configuration including defining the agent as a conversational interactive robot, designing a personalized logo, name, and introduction, and configuring knowledge base materials such as error research literature, Chinese proficiency level standards, and Chinese learner writing corpora. In the capability testing phase, the agent's basic abilities in multilingual communication, complex language expression, and contextual information understanding were tested, while its core capabilities were evaluated from four dimensions: accuracy, standardization, emotionality, and creativity. The test results showed that the agent performed excellently in multilingual recognition, Chinese proficiency, language standardization, emotional support, and creative thinking. In terms of comparative differences, multi-round testing revealed that the specialized agent, compared to general large models and other agents, is more targeted and professional in Chinese writing guidance, better meeting learners' needs.

5 Data Collection on Agent Usage and Evaluation Analysis Based on the CoI Theory Model

This chapter focuses on the application of the agent in International Chinese Language Education, exploring its role in writing instruction by collecting writing data from Thai international students and analyzing it based on the CoI theory model. The study selected five Thai international students at HSK5 level, collecting their assignment texts written in professional courses. After revision by the agent, a total of 61 rounds of dialogues and approximately 45,000 words of corpus data were generated. In terms of cognitive presence, the agent supports learners in knowledge construction through interaction, promoting higher-order cognitive engagement. It not only provides surface-level modifications such as grammar corrections but also involves cognitive integration and reflective processes, while demonstrating cross-cultural understanding, thereby enhancing learners' cognitive engagement. Social presence analysis shows that the agent enhances learners' emotional communication and socialization processes through emotional support and positive feedback, improving their sense of social presence and learning effectiveness, and promoting the construction of a human-machine collaborative and symbiotic learning community. In terms of teaching presence, the agent provides personalized instructional support by analyzing learners' needs and offering feedback based on the knowledge base, helping learners improve their language abilities and promoting cognitive and metacognitive development.

6 Prospects for the Development of the "Agent-Learner-Teacher" International Chinese Language Education Community in the Digital Intelligence Era

It explores the development of the International Chinese Language Education community composed of agents, learners, and teachers in the digital intelligence era. As a tool subject, agents enhance their professionalism through deep learning. In the future, they can be deployed locally based on open-source frameworks, optimize corpora, achieve intelligent writing assistant functions, provide personalized learning support for learners, and assist teachers in teaching management and intelligent grading. As learning subjects, learners' artificial intelligence literacy influences the effectiveness of human-computer interaction. Future research needs to focus on enhancing learners' literacy, critical thinking, and ethical awareness. As teaching subjects, teachers need to improve their artificial intelligence literacy, guide students in the correct use of technology, apply agents in teaching management and personalized instruction, and strengthen education on the ethics of technology use.

7 Conclusion

It summarizes the application achievements of generative artificial intelligence in International Chinese Language Education, emphasizes the application value of the agent constructed based on the ChatGLM model in enhancing the Chinese writing abilities of international students, and looks forward to future research directions, including improving the technical capabilities, professionalism, and specificity of the agent to promote the intelligent and personalized development of educational models, and to advance the digital reform and innovative development of International Chinese Language Education.