This IMLS-funded project aims to revolutionize how users engage with library and archival collections
Photo licensed by Adobe Stock
The AI-Collaboratory (AIC) has announced the launch of its latest initiative, HERITAGE–AI (Harnessing Enhanced Research & Instructional Technologies for Archival Generative Exploration using AI). Spearheaded by College of Information (INFO) Professor Richard Marciano, INFO PhD students Lori Perine and Rajesh Kumar Gnanasekaran, and Mark Conrad from AIC, this visionary project has secured a $194K IMLS National Leadership Grant to bring its ambitions to fruition.
The funding will support HERITAGE–AI’s inaugural project, GenAI-4-Arch (Harnessing Generative AI to Support Exploration and Discovery in Library and Archival Collections). This project aims to revolutionize how users engage with library and archival collections by integrating generative artificial intelligence (genAI) technologies. By partnering with the Maryland State Archives (MSA) and Kennard African American Cultural Heritage Center (KAACH), the initiative seeks to explore genAI’s potential while ensuring reliability and trust in its applications. Additionally, the team will develop methods to involve community members in the evaluation process, ensuring that the AI systems are responsive to the needs and perspectives of the communities they serve.
Central to GenAI-4-Arch is the development of a user-centered chatbot designed to enhance exploration and discovery within collections. This innovative tool promises to expand user access and provide more intuitive interfaces. Complementing the chatbot are additional activities, including community-driven AI design and evaluation, and a study of how to make archival collections AI-ready. By comparing AI-generated queries with traditional methods, the team aims to derive generalizable insights applicable to other cultural heritage collections.
The project’s approach will be rigorously tested and evaluated through focus groups, ensuring that outcomes such as improved access, enhanced user experiences, and comprehensive guidelines for AI use in libraries and archives are effectively realized. Ultimately, HERITAGE–AI seeks to strengthen the capabilities of libraries and archives, unlocking new dimensions of access, discovery, and understanding within their collections. The team plans to explore their research by expanding their focus on implementing specific generative AI technological aspects with agentic behavior and knowledge graphs to create more dynamic and intelligent systems for information retrieval. This approach aims to improve the efficiency of searches and uncover hidden connections within the archives, providing users with richer contextual information.
Perine shared her vision for the initiative, saying, “Our work will result in new insights for contextualization, participatory design, stakeholder evaluation, and metadata extensions for generative AI used with culturally sensitive data and related applications.”
As HERITAGE–AI forges ahead, the collaboration between academic, archival, and cultural institutions underscores a shared commitment to enriching public and research communities. By embracing cutting-edge technology and fostering inclusive design, the initiative is set to redefine how archives and libraries engage with the public, paving the way for a future where history and technology seamlessly intersect.