AI-Driven Data Sculptures to Promote Sustainability Content of Galleries, Libraries, Archives and Museums

Scientists predict dramatic consequences unless stronger actions are taken to reduce carbon emissions and mitigate climate change. The observable actions, however, are not commensurate with the threat. The reasons for this discrepancy are the lack of collective awareness (i.e., the perception of climate change as a geographically and temporally distant risk) and the lack of personal efficacy (i.e., the belief that their actions will not make a difference). Galleries, Libraries, Archives and Museums (GLAMs) are uniquely positioned to address this discrepancy through sustainability-focused collections and exhibitions. GreenGLAM will offer citizens and tourists alike new ways to identify and explore such offerings, enriched by data-driven immersive experiences. For this purpose, we introduce AI-driven “Virtual Data Sculptures” as a novel concept for visualising the sustainability content of GLAMs in mobile, Web-based and Augmented Reality (AR) environments. Such data sculptures will require the semantic processing of multiple content streams, including GLAM collection metadata and the public discussion (news, social media) about them. GreenGLAM will automatically extract factual and affective knowledge about GLAMs and their collections to visualise to citizens and tourists innovative insights into their relationship to the SDGs (Sustainable Development Goals) that will encourage further exploration and follow-up visits. Virtual Data Sculptures will help cultural organisations increase their visitors' sustainability knowledge, trigger co-creation processes and promote the transition to more sustainable lifestyle choices in line with the Sustainable Development Goals (SDGs). The resulting applications with these data sculptures can be disseminated and exploited by a wide range of cultural institutions. Professional stakeholders can also gain unprecedented transparency regarding the published content and activities of the GLAM sector. GreenGLAM will present an evolving window into the perceptions of sustainability content, revealing attitudes and opinions expressed in online media, both professional (journalistic) and public (social media). Such a rich feedback channel is crucial for GLAMs to further promote their collections and exhibitions and to attract new visitors - especially outside their traditional target groups. GreenGLAM will provide this integrated view and decision support to professional users via a Web-based visual analytics dashboard that integrates GLAM collection data and visitor interests, using deep learning models for language understanding and topic prediction. As such, GreenGLAM represents a novel data-centric approach to GLAM management, benefiting both visitors and the GLAMs themselves. The project will automatically classify aggregated GLAM content according to its relevance for the 17 SDGs. Virtual Data Sculptures will require radical innovation in scalable metadata enrichment, deep learning networks for language understanding and predictive analytics, coupled with advances in data visualisation and their integration into augmented reality applications. These innovations will reveal how past, current and future events and exhibitions relate to specific SDGs. This will not only help attract new target groups to cultural attractions but also show how to promote SDGs most effectively to increase awareness and nudge GLAMs and their visitors to reflect on and ultimately adopt more sustainable practices.