You are cordially invited to join the AI Empowering Special Collections of Libraries and the PRL Project, taking place on 30 June 2026 (HKT). This event marks PRRLA’s first co-hosted webinar with Fudan University, fostering collaboration among member institutions.
We warmly welcome all who are interested in AI applications in special collections to join this engaging and forward-looking discussion and share insights on advancing innovation in library services.
Event details:
- Date:
Monday 29th June 2026
6:30 – 8:00 PM (PST)
9:30 – 11:00 PM (EST)
Tuesday 30th June 2026
8:30 – 10:00 AM (WIB; THA)
9:30 – 11:00 AM (HKT, BST, SGT)
10:30 – 12:00 PM (JST, KST)
11:30 – 1:00 PM (AEDT)
- Speaker:
- (Moderator) Ms. Lijun Liu, Joan, Curator and graduate supervisor for the Master of Library Science (MLS) program at Fudan University Library
- Mr. Santiago Villegas-Ceballos, Founder of Medejean: Inteligencia Cultural and Colombian library innovation leader
- Mr. Haiqing Lin, Head of Technical Services, C. V. Starr East Asia Library, University of California, Berkeley
- Ms. Song Xue, Associate Research Librarian and Director of Data Management and Technology Department, Fudan University Library
- Agenda:
Welcome and Introduction
Sharing Session:
Interactive Q&A Session
Closing Remarks
We warmly encourage you to attend and share this invitation with your colleagues.
For inquiries, please contact us via prrla-sec@cuhk.edu.hk. We look forward to your participation and to making this event a great success together!
Biography:
- (Moderator) Joan Lijun Liu is a research librarian at the Institute of Humanities and Social Science Data and concurrently a graduate supervisor for the MLS program at Fudan University Library. Prior to her current appointments, she taught law in China and later worked as a tenured associate curator at the NYU Law Library. Her work has long focused on the intersection of library and information science and law, particularly in the areas of legal information access and legal knowledge organization. She was formerly an active member of ALA and AALL, and as a long-standing member of IFLA, she currently serves on the Standing Committee of the Law Libraries Section.
- Santiago Villegas-Ceballos is a Colombian library innovation leader and founder of Medejean: Inteligencia Cultural. He leads the Innovation and AI Unit at eLibro, co-directs IAparaBibliotecas.com, and serves as an elected member of the IFLA Information Technology Section Committee. With more than two decades of experience in human-centred digital transformation, he has supported libraries and cultural institutions across four continents. His work focuses on responsible AI, critical information literacy, and the design of inclusive library services from a Global South perspective.
- Haiqing Lin is serving as the Head of Technical Services at the East Asian Library, University of California, Berkeley. Before joining Berkeley, he worked as the East Asian Studies Librarian at the University of Auckland Library in New Zealand. His research interests focus on the application of modern technology in library services. Currently, He is exploring how to leverage AI technology to generate metadata for special collections to enhance the accessibility of these unique materials.
- Xue Song, Associate Research Librarian at Fudan University Library. She serves as Director of the Data Management and Technology Department and Deputy Director of the Fudan-SH RFID Joint Research Center for Smart Library. With extensive experience in information systems development, data management, and technology applications in academic libraries, she has contributed to key university projects, including the Fudan University Smart Library Digital Base and the Fudan University Library Service Platform. Her research focuses on smart library platforms, library data governance and utilization, and the application of intelligent technologies in academic library services.
You are welcome to rewatch the recording of the session here.
Q&A-Response from Ms. Xue Song:
Is the eLRM data model still stored in MARC format in the backend database?
No, not in MARC format. For the data organized by the eLRM model, we store it in our library data center. The backend database stores entities, attributes, and relationships. So MARC records can be one of the original data sources, but after mapping and transformation, the integrated data is no longer stored as MARC. At this stage, we are not using a graph database or RDF triples. It is still stored in a relational or structured database, but the data structure is designed around entities and relationships.
Is the OCR text preprocessed and stored in advance, or is it generated concurrently/on demand? How many forms or representations of the content do you maintain in storage?
The OCR text is preprocessed and stored in advance. For page-level display, the system does not generate OCR text on demand. It reads the processed OCR text from storage and displays it together with the page image. For preservation and reuse, we maintain several representations of the content: the original scanned files, the OCR text, the translated text, and related metadata. We also allow multiple versions of OCR text and translations to be stored, because the results may be revised after human review or improved when better models and tools become available.