News

The goal of MODAL is to explore the potential applications of generative artificial intelligence (GenAI) for the cultural heritage sector and to disseminate this knowledge within the sector. In October 2024, ADVN, in collaboration with Amsab-ISG, Letterenhuis, MoMu, and meemoo, launched the project "Metadata and Access to Digital Archives using Large Language Models," or MODAL for short. Large Language Models (LLMs) are large-scale language models trained to generate text by processing immense amounts of textual data. The goal of MODAL is to explore the potential of this form of generative artificial intelligence (GenAI) for the cultural heritage sector and to actively share this knowledge within the sector. Cultural heritage is increasingly taking on digital forms. The exponential influx of born-digital archives poses a huge challenge for archivists and collection managers in terms of management, processing, and access. As a result, many digital archives and collections remain inaccessible and invisible to researchers and the general public. The advent of GenAI creates new opportunities to unlock heritage more efficiently and richly. Using LLMs, collection managers and archivists could process unstructured textual data more quickly—for example, by creating summaries, classifying texts by form, or generating keywords. The ability to search extensive collections more user-friendly and effectively could offer groundbreaking opportunities for diverse users. Within MODAL, three use cases will be explored: making digital archives searchable, detecting sensitive or problematic content, and making audio material textually searchable. Based on these three use cases, MODAL will not only develop and test several tools but also investigate how these tools can be implemented into the workflows of archivists and collection managers. Since January 2025, Datable, a consultancy specializing in digital heritage, has been working with research material provided by the various project partners. The MODAL project aims to broadly share research results and stimulate collaboration and knowledge exchange regarding the use of LLMs and GenAI in the cultural heritage sector. This will be achieved by deploying the developed scripts as deployable applications for the sector and by establishing a collaborative research group.