Exhibition online content with generated viewing paths
The Sapporo International Art Festival (SIAF) is an art festival held in Sapporo once every three years. In addition to introducing contemporary art and media art, the festival promotes the attractions of Sapporo both in Japan and abroad. The third SIAF2020 was canceled in consideration of the impact of the spread of the COVID-19, and a new website called “SIAF2020 Matrix” was released as online alternative. Qosmo was in charge of the planning and production of this website in collaboration with Agnieszka Kubicka-Dzieduszycka, the planning director of SIAF2020, and the secretariat of the SIAF2020 Executive Committee.
Between 2020 and 2021, pandemics will permeate our lives, and various regulations will make it difficult for people to gather at real events to share spatial experiences. Under these circumstances, the Sapporo International Art Festival (SIAF2020) team came up with the idea of designing and realizing a new online platform instead of a real event, and the project was launched. In the course of planning, we reconsidered what an exhibition space is, and focused on building a living experience of the original exhibition that was to be held, and reviving it in an online space. As a result, we have created a website that is not just an archive of artworks, but also an online platform that changes according to the interests of the viewer.
This website uses machine learning to generate a unique viewing path based on the keywords chosen by the visitor and allows the visitor to follow the path to view each artwork.
First, the layout displayed on the home page was designed to be a reflection of the actual coordinates of each work, originally placed around the city of Sapporo, Japan — these locations were an intentional choice, creating a curated map of the city to be followed by visitors. In our virtual projection, the webpage opens with this geographical context as a background, to then fade and evolve. As if they were trying to adjust to this new medium, the artworks dynamically re-organize themselves to fill the computer screen, following the rules of physics and become objects with which users can interact.
As visitors select and proceed into the exhibition, the angles formed by the current path and the following are visible on the screen, like corners and corridors. This reminds the visitor of the spatial context without degrading the navigation function.
We decided to provide a way for visitors to get a sense of proximity between them, similar to a physical room. In order to achieve this, anyone viewing artwork is being anonymously sent in real-time to the server, to then be dispatched to the others. As you view the main exhibition layout, the presence of those “around” you are brought to your attention, as they materialize on your screen.
Beneath the surface, we brought in the modern computing techniques, machine learning that we often make use of at Qosmo. The idea is to build a generative narrative: using the available data of each artwork, we wanted to propose a guided visit that would reflect the meaning of the artworks. Here, we decided to focus on the concepts, carefully written for each of them. When put through natural language processing models, features can be extracted from these texts. It suggests to the visitor a viewing path that reflects the meaning of that artwork. While this space by default is made of a high number of dimensions, it is in our case reduced to just two, for performance reasons: those two dimensions are essentially similar to those of a map, where proximity is a straightforward distance calculation.
The above is where the navigation of the website takes place. Starting from any random project, the user can be guided to the closest one, and so on. Thus, using modern data processing tools, we bypass the traditional arbitrary web navigation and propose a new, natural way to consume information.
Agnieszka Kubicka-Dzieduszycka (SIAF2020 Curatorial Director of Media Art）
Sapporo International Art Festival Executive Committee
Robin Jungers (Qosmo, Inc.)
Shoya Dozono (Qosmo, Inc.)
Naoki Ise (Qosmo, Inc.)
Sakiko Yasue(Qosmo, Inc.)
Nao Tokui (Qosmo, Inc.)
Taro Amano (Curatorial Director of Contemporary Art and Director in Chief of SIAF2020)
Asami Hosokawa (SIAF2020 Secretariat Manager of SIAF2020)
Naoto Iwasaki (Curator of of SIAF2020 (Hongo Shin Memorial Museum of Sculpture, Sapporo))
Kazumi Miyai (Curator of SIAF2020 (Moerenuma Park))
Emi Nakamura (Curatorial Advisor of SIAF2020)
Seiji Nakamura (Curator of SIAF2020 (Hokkaido Museum of Modern Art))
Tsubasa Nishi (SIAF2020 Secretariat Manager of SIAF2020)
Kazuhiko Sakurai (Director of Hobetsu Museum)