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).