Interpretable Clusters for Representing Citizens’ Sense of Belonging through Interaction with Cultural Heritage

Abstract

The EU H2020 project SPICE (Social cohesion, Participation, and Inclusion through Cultural Engagement) focuses on developing, designing, and implementing new methods and digital tools for citizen curation. This paper delineates several software tools developed within the project, presenting innovative approaches to represent and visualize citizens and communities resulting from their engagement with cultural heritage. Aligned with the central tenets of SPICE –particularly the notions of belonging and the Interpretation Reflection loop– the primary objective is to bolster citizens’ participation and inclusion in fostering social cohesion. This paper describes how the SPICE tools can be utilized to guide the processes of interpretation and reflection on cultural heritage artefacts. The Community Model serves as a pivotal component, enabling the modeling of citizens and communities through the utilization of similarity functions for clustering citizens based on perspectives. The clustering algorithm is intricately crafted to generate coherent communities, iterating until all clusters are interpretable using demographic attributes, centroid-based representations, and similarity attributes. Authors posit that this model holds significant value in comprehending and structuring complex data within cultural heritage contexts. To exemplify our approach, the paper examines different attributes of individual citizens and citizen groups in the GAM (Galleria Civica d’Arte Moderna e Contemporanea) case study. Here, perspectives are delineated based on visitors’ demographic attributes and their emotional responses when engaging with artworks. These perspectives are then visualized using the VISIR tool, facilitating the exploration and revelation of connections between citizens and communities, thereby bridging the realms of citizen space and cultural heritage space.