Background

In recent years, there has been significant progress in the digitalization of cultural artifacts and collections. The spectrum includes everything from archival material like baptismal registrations and cadastres to works of art like photographs or amateur videos, all the way up to entire city districts. To make these vast collections navigable to both experts and the broader public, user-friendly interfaces are indispensable. User-friendliness necessitates straightforward, yet widely applicable data visualizations that aid users in effectively accessing data, finding patterns and relationships, and making conclusions. Moreover, automatic analysis methods are required to capture the complex and diversified content.

Project Content

We investigate the usage of interactive applications for two scenarios in this collaborative PhD program: In scenario one, we look at how automated, AI-assisted content analysis and expert knowledge can complement each other, which unique insights emerge from this, and how these insights can be converted into practical outcomes. In scenario two, we use interactive narratives (storytelling) to share our learnings with the general audience.

In two aspects, the project is interdisciplinary. On the one hand, it builds a bridge between computer science and the humanities. On the other hand, the computer science disciplines Visual Analytics and Computer Vision/Machine Learning deepen their collaboration. The two participating institutions (St. Pölten University of Applied Sciences and TU Wien) each provide around half of the doctoral program's faculty members. The doctoral students can carry out research at both institutions, learn more about all departments involved, and establish valuable contacts for future collaborations.

Goals and work steps

The overarching goals are worked out in five dissertation projects that can be split into three conceptual levels: (1) metadata generation, (2) data exploration, and (3) crafting data experiences. The research focuses on historical archival collections, particularly in the fields of photography and film, while also exploring new approaches and methods, including:

• Automated, content-based analysis of historical film footage.
• Automated techniques for identifying the camera and lens types used in historical photographs. These techniques rely on the photographic material itself as well as on what is known about the photographic material (such as location, period, and historical classification).
• Exploratory analysis of historical image collections with little or no metadata.
• Using visual analysis to analyse cultural asset networks. Spatial and temporal links between cultural assets are revealed, allowing for new perspectives on historical events and situations.
• Creating new techniques of situational storytelling to win over a broader audience for the project's content. This involves connecting digital content with physical spaces and tailoring it to the interests and actions of the users.