HathiTrust Receives $5M Grant for AI Book Discovery Tools
The $5 million Mellon Foundation award funds a five-year project to improve search and metadata for 19 million digitized books.
Key points
- HathiTrust received a $5 million Mellon Foundation grant for a five-year AI project.
- The initiative aims to improve metadata and search for 19 million digitized books.
- Partners will develop ethical guidelines to ensure noncommercial and transparent AI use.
HathiTrust has secured a $5 million grant from the Mellon Foundation to deploy artificial intelligence across its massive digital library. The funding supports a five-year initiative designed to make it easier for researchers and the public to find and use information within the organization’s collection of 19 million digitized volumes.
Reported by Research Information, the project aims to build shared, academic-led infrastructure. This approach intends to balance advanced technology with strict adherence to legal and ethical standards, ensuring that digitized materials remain accessible for noncommercial scholarly use.
Improving search and metadata quality
A primary goal of the initiative is to enhance the descriptive metadata and text quality of existing books. By using AI to better identify subjects and languages, the system can generate more accurate summaries and create interactive search tools. This should allow users to explore specific books or broader collections with greater precision and ease.
The project also seeks to provide verified sources of book text and images at scale. This infrastructure could support advanced computational analysis and text mining, potentially enabling the development of specialized language models for academic purposes.
Establishing ethical and legal frameworks
Technical improvements are accompanied by a focus on governance. HathiTrust is working with Northeastern University Library and Authors Alliance to develop shared principles covering legal frameworks and research integrity. These guidelines are intended to ensure that accountability and the public interest remain central to how the technology is deployed.
The partners also plan to address the environmental impact of AI processing. By defining these standards early, the project aims to prevent the unchecked expansion of AI tools in academic settings, ensuring that scholarly values are not compromised by commercial pressures.
Challenges in sustainability and access
A significant trade-off in this project is the need for financial sustainability. The team must determine how to fund these services long-term without introducing commercial models that could restrict access. This is a critical challenge, as the goal is to keep the resources free for academic and public-interest uses.
HathiTrust executive director Mike Furlough emphasized that libraries have historically maintained their values despite rapid technological change. He stated that this project continues a longstanding effort to make the scholarly record more accessible, ensuring that the interests of publishers, authors, and the public are all considered in the process.
