Speaker
Description
Urban and territorial systems face increasing challenges due to population growth, climate change, and social inequities, necessitating transformative approaches to spatial planning.
Urban green infrastructure, as a socio-ecological connections (Davies et al., 2006), addresses these challenges driving ecological, climatic, social, and cultural transformation. Among its benefits, Cultural Ecosystem Services (CES) emerge as key interpretative and operational tools guiding decision-making towards more resilient, sustainable, and community-responsive cities (Fish et al., 2016).
Traditional CES assessment methods, subject on time-intensive surveys, limit scalability and frequency of updates. By contrast, social media and online reviews provide extensive, readily accessible datasets that enable continuous and specific assessments. Emerging digital technologies, particularly artificial intelligence (AI) and crowdsourced data, offer new opportunities and perspectives for mapping and assessing CES, and therefore inform and guide planning processes shaping resilient, inclusive urban systems. This study capitalizes on these advantages and presents an innovative model that combines GIS-based mapping, generative AI, geolocated crowdsourced data from Google Maps reviews to analyze, evaluate and map CES offering insights into the social dimension of urban green infrastructure.
The model operates in four key stages: (i) definition of strategic CES and the analytical protocol, (ii) acquisition and catalog of Google Maps data, (iii) evaluation, and (iv) spatial mapping and visualization. The first phase identifies the strategic CES provided by urban green infrastructure and defines indicators for their assessment. Based on Fish et al. (2016), the structured analytical protocol focuses on human-environment “interactions” as descriptors and enablers of CES. The four CES categories analyzed are: (i) Physical recreation, (ii) Social relations and cohesion, (iii) Education, learning, and inspiration, and (iv) Aesthetic and cultural heritage value. The second phase involves acquiring geolocated textual reviews from Google Maps and spatial data from OpenStreetMap. This initial spatial analysis establishes a baseline for evaluating CES. Next, geolocated reviews from Google Maps are acquired daily through a custom-built application, which processes user-generated data (places, reviews, ratings) and stores it in a centralized database. In the third phase, Google Maps data is processed using a generative AI Deep Learning LLM (Large Language Model) algorithm. A custom-built, web-based front-end panel allows us to query this archive using Natural Language Processing (NLP), providing maximum flexibility in formulating questions to extract the necessary information. With this deep learning tool, reviews are automatically processed and analyzed to link them to specific CES and quantify their benefits. The final phase integrates processed data into a geodatabase to generate thematic maps illustrating: (i) overall CES distribution, (ii) Physical Recreation, (iii) Social Relations and Cohesion, (iv) Education, Learning, and Inspiration, (v) Aesthetic and Cultural Heritage Value, (vi) CES co-occurrence, and (vii) aggregated popularity indices. These outputs facilitate an exploration of community perceptions, spatial variability in CES provisioning, multifunctionality of urban green infrastructure and critical planning metrics, allowing planners to identify user preferences, sentiments, and demands.
The application of the in the context of Rome demonstrates its potential to support transformative planning providing dynamic, data-driven insights. The model is operational, scalable, and replicable, leveraging AI, digital tools, and temporally defined georeferenced data to ensure comprehensive CES evaluation. This ensures the mapping and evaluation of the complexity of human-nature interactions from both quantitative and qualitative perspectives. It is not only spatially specific but also repeatable and monitorable over time, positioning it as a system to support urban-scale analysis, planning, decision-making, and monitoring processes, with a focus on the social dimensions of urban green infrastructure within sustainable urban regeneration scenarios.
References
Davies, C., McGloin, C., MacFarlane, R. & Roe, M. (2006) Green Infrastructure Planning Guide Project: Final Report. NECF.
Fish, R., Church, A. & Winter, M. (2016) Conceptualising cultural ecosystem services: A novel framework for research and critical engagement Ecosystem Services, 21(B), pp. 208-217.
| Keywords | cultural ecosystem services; urban green infrastructure; urban planning; artificial intelligence |
|---|---|
| Best Congress Paper Award | Yes |
