AI Strategy & Projects

ULS Strategy

Utilize existing AI tools to improve access to information, enhance existing services, or as the basis of new services.
Identify and partner with researchers on the development of tools, models, etc. where it makes sense and where mutual benefit exists.
Support campus AI literacy by offering classes and workshops.
Foster an AI literate ULS acknowledging that different roles will need different areas and levels of expertise (e.g., subject experts develop expertise in AI tools in their discipline(s)).
Be a source of data (full-text, structured data, etc.) from both library-owned and library-leased sources for use in campus AI and ML initiatives.

ULS Projects

The library was involved in an NEH grant proposal, "Linking voices across time: a federated AI approach to connecting Japanese visual cultural heritage collections." Partners included Rutgers University and Durham University.

Using Transkribus to read and transcribe handwritten archival content

Will be investigating the use of Annif, a tool for automated subject indexing and classification

Collections assessment project: The project aims to use ChatGPT to assign LC classes to research and teaching activities (publications, courses taught, and grants applied for) at Pitt. These will be compared against the current ULS collecting profile (e.g., Gobi and other sources) and the overall collection profile to identify areas of strengths and gaps.