Narratives from the long tail

Transforming access to audiovisual archives

Bridging academia, archives, museums and society, Narratives from the long tail takes up the contemporary challenges of public access to the principal mnemonic records of the 20th and 21st centuries: large-scale audiovisual archives. Through computational processes, Narratives sets out to address and resolve the gap between digital archives and the embodied, participatory world of museological experience. This interdisciplinary innovation will be led by four exemplary academics in machine learning, visual analytics, digital museology, and archival science. Taking a systems thinking approach to incorporate  all aspects of its dynamic structure, together we will pioneer ‘computational museology’ through interlocking methods that will allow audiences to meaningfully explore the semantically rich ‘long tail’ of audiovisual memory. This project is a collaboration between the Laboratory for Experimental Museology (eM+), EPFL; Visual intelligence for Transportation Laboratory (VITA) EPFL, Visualization and MultiMedia Laboratory (VMML) University of Zürich and; Faculty of Humanities. University of Amsterdam (UvA).

Switzerland has made exemplary investments in the digitization and curation of audiovisual cultural heritage; however, outside of domain specialists, the vast majority of these collections remain inaccessible to broader society. To address this grave shortcoming, Narratives transcends state-of-the-art across all its fields to initiate a groundbreaking visualization framework for interactively (re)discovering hundreds of thousands of hours of audiovisual materials. This is achieved through Narratives’ highly novel 360-degree 3D Narrative Visualization Engine, that will dynamically reveal the spatial, temporal, social, affective and aesthetic patterns, phenomena and processes of archival collections, and ensure ‘narrative coherence’ for audiences as they navigate them. This scalable and participatory engagement model aims to situate audiences and their experiences at the center of knowledge production while also stimulating innovation in archival practices.

For this purpose, we will set a computational agenda for innovation through our complimentary consortium, capable of integrating, analyzing, operationalizing, visualizing and sharing these audiovisual archives, grounded in methodological and theoretical criticality. The transversal methods that we will employ have the potential to exceed current approaches to audiovisual archives through three interlocking frameworks: Firstly, by systematizing machine learning and visual analytics as primary operations within the data curation pipeline, we will design and test methods to augment conventional archive data leading to abundantly richer semantic descriptions for these collections. Secondly, alongside rigorous evaluation, we will design a narrative-seeking visualization framework, combining spatio-temporal, social, aesthetic and affective semantics to pioneer new audiovisual storytelling formats. Thirdly, by creating a scalable visualization architecture that can be applied to other collections and diverse interfaces (e.g. 360-degree; fulldome; tiled and head mounted displays), we will significantly advance on current practices to develop platforms for the navigation, exploration and creative reorganization of moving images through immersive and interactive systems.

Narratives draws on leading digital archives to provide the basis for its research methods including: Radio Télévision Suisse, the UNESCO Memory of the World listed Montreux Jazz Festival archive, and the Netherland’s Eye Filmmuseum’s Mutoscope and Biograph collection, in Amsterdam. Chosen for their qualitative global significance and diversity in form, content and volume, these representative collections are straining at the boundaries of traditional archiving and museological processes and are demanding new solutions from those who want to engage with them. In addition, for copyright reasons alone, these like most audiovisual archives, are limited to in situ viewing, putting further strains on sustainability. These serious limitations stimulating Narratives to innovate gallery-based solutions with globally applicable outcomes.

Through a framework of computational museology, Narratives initiates transformations in cultural transmission though narrative visualization. It harnesses epistemological rigor pursuant to curatorial and archival theory to ensure unified high-impact outcomes in science, humanities, cultural industries and society in the form of books, articles, symposia, software and installations. Through its model of shared value together with memory institutions and their audiences, Narratives advances the Swiss Federal Council’s (2021-2024) strategic cultural policies of cultural participation, social cohesion, innovation and digital change while it also validates many years of effort in digital preservation and curation that has taken place across Switzerland, to ensure the sustainability of our collective heritage.