Semantic Technologies for Global Governance: A Hybrid AI Approach to Tracking and Monitoring WHO Resolutions

Deliberative institutions, such as United Nations agencies, face an urgent need to track and monitor the decisions and obligations articulated in their regulative acts. To address this challenge, we present a comprehensive software architecture that combines Artificial Intelligence (AI), ontologies, and knowledge extraction methodologies to support the detection, extraction, annotation, and longitudinal monitoring of provisions in World Health Organisation (WHO) resolutions. The proposed system integrates a web-based user interface, a knowledge extraction module, and a knowledge graph to formalise and interlink data, enabling the identification of deliberative components such as actors, roles, decisions, and events. This knowledge is used to generate the agendas for governing body meetings, including the WHO Executive Board and the World Health Assembly. By employing a hybrid AI approach rooted in semantic technologies and established ontology design patterns, the system ensures robust representation of mandate provisions and their relationships to other documents and events. Additionally, the user interface captures expert input to refine extracted knowledge and improve system accuracy. The proposed methodology aims at addressing critical needs for avoiding repetition, missed reporting, and redundant deliberations in global governance.