A Statistical Framework for Detecting Emergent Narratives in Longitudinal Text Corpora
Published in arXiv (Working Paper/Preprint), 2026
This paper proposes a statistical framework for detecting and characterising the emergence of narratives in longitudinal text corpora. Using probabilistic topic models in combination with lag-correlation analysis, we develop methods for identifying when topics acquire narrative coherence and tracking their diffusion across time.
Recommended citation: Medeiros, C., Quigley, J., & Revie, M. (2026). A statistical framework for detecting emergent narratives in longitudinal text corpora. arXiv:2602.20939.
Download Paper