With the emergence of DeepSeek and the rapid development of AI generative search, our university library has implemented "AI Generative Resource Search" based on its academic literature collections. It features the following aspects:

1.‌Content Generation Based on Library Resources‌: The generated content is traceable to the original literature, enhancing accuracy and reliability.

2.‌Smooth Generation Based on DeepSeek‌: The content is smoothly generated and highly relevant to the query.

3.‌Citation Functionality‌: It enables one-click copying and downloading of references in multiple formats.

4.‌Related Resources Display‌: The generated page showcases related library resources, including books, journals, and the university's research achievements.

Over the past six months, the library has held 5 seminars and 7 system debugging meetings, comparing the efficiency of different models and completing page design. After 3 rounds of user experience testing with faculty and students, which garnered their approval, the service is officially launched from today.