Generative Artificial Intelligence Empowers the Research of Digital Textbooks in Vocational Education

This research focuses on the application of AIGC in vocational education digital textbooks and how RAG technology can be used to further improve the interactivity and educational effectiveness of digital textbooks. The paper first analyzes the development of AIGC technology and its great potential in the field of education, especially the role of LLM in content generation. It then discusses the concept, risks, and construction strategies of vocational education digital textbooks, emphasizing the importance of digital textbooks in teaching and their differences compared with paper textbooks. The study also designs a knowledge retrieval question-answering system based on RAG, which combines school-based textbooks and cooperative enterprise manuals as a knowledge base to generate responses to student questions. While ensuring accuracy and completeness, the quality of the responses was improved by the merge of students' queries with relevant texts using RAG technology. Regarding implementation, RAG technology consists of knowledge base construction and knowledge retrieval. It includes segmenting the document, recognizing and vectorizing it in order to build a vector database. During user retrieval, queries are sorted through the vector database interface, and the matched knowledge fragments and query statements are merged to generate answers including specific domain knowledge. This framework of the RAG question-answering system can provide accurate answers to satisfy students of all levels in personalized teaching while maintaining explainability and timeliness of the content.