In working closely with our user community, Ex Libris, a part of Clarivate, understands the challenge of compiling complete metadata records at scale.  

With diverse content types in numerous formats coming from varied sources, maintaining consistency and connectivity in metadata has become more important – and more complex – than ever before.  

Until now, catalogers and library staff have dedicated significant time and effort into creating robust metadata records to simplify cataloging on the backend, and to improve searchability for users on the frontend.  

Richer records make more comprehensive work of collection development, collection management, and content discovery – but manual generation of intensive Bibliographic data is a Sisyphean task. It is nearly humanly impossible to input complete descriptive information for entire library collections, which are constantly in flux.  

So, we asked, “what if it was possible to recruit AI to help?” 

Introducing the AI metadata generator from Ex Libris  

We are excited to share the creation of an AI-driven metadata generator that has already been able to enrich Alma Community Zone Bibliographic records. The current enrichment is focused on AI generated metadata for three specific MARC fields:  

• Language (041 and 008) 
• Summary (520) 
• LC subject headings (650)  

We began the development of this revolutionary technology with the repository of MARC records in the Alma Community Zone, by letting AI process the full-text or partial full-text of a book, and then automatically generate metadata based on the content. Our focus centered on the above fields of language, summary, and subject headings in alignment with the Library of Congress standards.  

As of February 2024, our AI metadata generator is live with 200 ebook titles from ProQuest EBook Central – and counting. There is a lot of work left to do as we continue adding more titles, open access to others and deepen our AI generated metadata collaboration with ProQuest resources and other Content Providers.  

For more details and updates on our evolving status of AI generated metadata, read the article in the Ex Libris Knowledge Center.  

Why is this important? 

This scalable technology will have a profound impact on catalogers, library staff and their users. AI generated metadata enrichment will not only save library staff time and effort; it ensures Bibliographic records encompass comprehensive – and therefore useful – information. For collection development and collection management purposes, this means gaps in available resources can easily be surfaced, and new resources from various sources can be effortlessly incorporated into the catalog. 

Additionally, AI generated metadata enhancements provide more robust Bibliographic information that will make collections more discoverable to library users. This will support users in accessing relevant resources that support their learning goals and research areas, opening the way to exciting and often novel insights, knowledge and understanding – as well as potential increases in usage of library resources.  

 At Ex Libris, the use of AI technology is always done to solve the real challenges of real users, librarians, and libraries. We respect the need for privacy and intellectual property considerations, and above all, we understand the importance of using trusted scholarly content to generate results and information.    

This article is the fourth in our series on AI developments at Ex Libris. Read our previous blog posts on A Unified Approach to Title Matching in Libraries.    

Academic libraries are vital for accessing trustworthy scholarly material. Download this whitepaper to learn how new methods and technologies, like Generative AI, can enhance the library’s discovery experience and align it with dynamic user expectations.