Graph-based LLM over Semi-Structured Population Data for Dynamic Policy Response
Daqian Shi, Xiaolei Diao, Jinge Wu, Honghan Wu, Xiongfeng Tang, Felix Naughton, Paulina Bondaronek
TL;DR
The paper tackles the challenge of timely interpretation of semi-structured population data during public health emergencies by introducing a graph-based LLM framework that links structured demographic attributes with unstructured feedback via a dynamic, need-aware graph. It presents a three-module pipeline—data pre-processing, needs extraction, and need-aware analysis with visualization—that enables weak supervision, online graph enrichment, and interpretable insights for policy decisions. Demonstrated on a real-world COVID-19 UK dataset (1,045 participants, 3,812 text responses over 24 months), the approach reveals temporal shifts in needs and demographic heterogeneity, including sentiment dynamics. By integrating a Mechanism of Action ontology mapping and LDA-derived topics within graph-constrained LLM reasoning, the method offers scalable, interpretable population monitoring suitable for resource-constrained health systems and rapid policy adaptation.
Abstract
Timely and accurate analysis of population-level data is crucial for effective decision-making during public health emergencies such as the COVID-19 pandemic. However, the massive input of semi-structured data, including structured demographic information and unstructured human feedback, poses significant challenges to conventional analysis methods. Manual expert-driven assessments, though accurate, are inefficient, while standard NLP pipelines often require large task-specific labeled datasets and struggle with generalization across diverse domains. To address these challenges, we propose a novel graph-based reasoning framework that integrates large language models with structured demographic attributes and unstructured public feedback in a weakly supervised pipeline. The proposed approach dynamically models evolving citizen needs into a need-aware graph, enabling population-specific analyses based on key features such as age, gender, and the Index of Multiple Deprivation. It generates interpretable insights to inform responsive health policy decision-making. We test our method using a real-world dataset, and preliminary experimental results demonstrate its feasibility. This approach offers a scalable solution for intelligent population health monitoring in resource-constrained clinical and governmental settings.