Abstract

The global challenge of food insecurity poses a significant obstacle to achieving Sustainable Development Goal (SDG) 2: Zero Hunger. Tracking this problem is a complex task, as food security is influenced by a multitude of interconnected factors —from climate change and agricultural productivity to geopolitical conflicts and socio-economic policies. In this context, timely and accurate monitoring has become imperative. To address this challenge, the FAO’s Data Lab proposes an innovative approach for obtaining early insights into global food insecurity trends. Our methodology harnesses artificial intelligence (AI) to classify news articles related to food insecurity, coupled with a neural network algorithm to gauge the sentiment. This enables real-time assessments of the evolution of food insecurity across diverse regions worldwide. Through extensive evaluation using a broad corpus of news articles spanning diverse geopolitical context and languages, our system demonstrates a robust capability to accurately detect and flag instances of worsening food insecurity.
The proposed system positions itself at the forefront of integrating cutting-edge technology with vital humanitarian needs, where the swift identification of negative trends in food security through media sentiment allows for preemptive actions and policy interventions. By exploiting the power of AI to analyze vast amounts of news articles, our approach offers a scalable and responsive tool to support the global fight against hunger, directly aligning with and enabling progress toward the Sustainable Development Goals.