Accelerating biodiversity restoration and conservation with AI-driven evidence synthesis (HYBRID)
by
Abstract:
Ecosystem degradation driven by human activities and climate change has caused major losses in biodiversity, ecosystem functions, and services worldwide. Global initiatives such as the UN Decade on Ecosystem Restoration (2021–2030) and the Convention on Biological Diversity’s Target 2 call for the effective restoration of at least 30% of degraded ecosystems by 2030. Ecological restoration is an important tool to halt and reverse the collapse of biodiversity decline, but its practice largely relies on ad hoc decisions by busy practitioners and policymakers. Semantic knowledge graphs have the capacity to represent knowledge from both the peer-reviewed scientific literature and experiential knowledge. We are developing a knowledge graph using grassland restoration studies that will be enriched with practitioner interviews and paired with a LLM to make a public-facing tool for grassland restoration. We will test this tool and evaluate its accuracy, trustworthiness and efficacy with a community of practice. I will share more details about the architecture, key decisions and opportunities for research collaboration.
Speaker Bio:
Tim is a scholar working on engagement, program participation and open knowledge systems in biodiversity restoration and conservation. He is a postdoctoral fellow with the Bennett Lab at Carleton University (Canada), an Adjunct Associate Professor in the School of Environment, Resources and Sustainability at the University of Waterloo (Canada) and a researcher with Waterloo.AI.
Zoom:
https://tu-dortmund.zoom.us/j/92058513225?pwd=DHa5Zk9Z3bewRH5mjkI3Aan98KQ8JJ.1
Dr. Raphael Fischer
Dr. Jens Buß