BEYOND's work on AI-based early warning systems presented in China

On 23 Nov 2025, Claire Nédellec presented “Leveraging NLP and LLMs for AI-Based Phytosanitary Early Warning Systems” for the Conference on Agricultural Knowledge Engineering AKIE25 organized by the Digital Agriculture Committee of the China Computer Federation (CCF) and held at Huazhong University in Wuhan, China. 

Abstract: Natural Language Processing (NLP) is emerging as a key enabler of AI-based early warning systems for crop pest and disease surveillance. This talk presents our work on on structured information extraction in plant health, covering entity recognition, taxonomic and geographical normalization, and complex epidemiological event modeling. I will introduce the EPOP corpus, a novel multilingual resource we release for benchmarking NLP methods in this domain. Finally, I will share results from our experiments using both fine-tuned domain-specific models and prompted large language models (LLMs). These advances aim to support timely, data-driven responses in phytosanitary monitoring and foster research in environmental NLP, crop protection, and knowledge base enrichment.

See the presentation here.