Harnessing artificial intelligence technology and social media data to support Cultural Ecosystem Service assessments

A user-friendly AI-based approach, combining computer vision and text mining, successfully links photographers' preferences from 640,000 Flickr images to four key Cultural Ecosystem Services (CES) categories—aesthetic value, outdoor recreation, cultural heritage, and symbolic species—with over 90% accuracy. The study demonstrates that these preferences can be spatially predicted using environmental and infrastructural variables, revealing CES hotspots in high-tourism and residential areas of the Dolomites UNESCO World Heritage Site. This method offers a rapid, flexible, and transferable alternative to traditional resource-intensive assessments, providing valuable insights for sustainable environmental management.

Egarter Vigl, L., Marsoner, T., Giombini, V., Pecher, C., Simion, H., Stemle, E., Tasser, E., & Depellegrin, D. (2021). Harnessing artificial intelligence technology and social media data to support Cultural Ecosystem Service assessments. People and Nature, 3(3), 673–685. Portico. https://doi.org/10.1002/pan3.10199