Reflective thinking meets artificial intelligence: Synthesizing sustainability transition knowledge in left-behind mountain regions

Abstract

We demonstrate a multi-method approach towards discovering and structuring sustainability transition knowledge in marginalized mountain regions. By employing reflective thinking, artificial intelligence (AI)-powered text summarization and text mining, we synthesize experts’ narratives on sustainable development challenges and solutions in Kardüz Upland, Türkiye. We then analyze their alignment with the UN Sustainable Development Goals (SDGs) using document embedding. Investment in infrastructure, education, and resilient socio-ecological systems emerged as priority sectors to combat poor infrastructure, geographic isolation, climate change, poverty, depopulation, unemployment, low education levels, and inadequate social services. The narratives were closest in substance to SDG 1, 3, and 11. Social dimensions of sustainability were more pronounced than environmental dimensions. The presented approach supports policymakers in organizing loosely structured sustainability transition knowledge and fragmented data corpora, while also advancing AI applications for designing and planning sustainable development policies at the regional level.