A new arXiv preprint argues that existing evaluations of topic models for short social media texts rely almost entirely on measures of the generated topics themselves. The author proposes metrics that also assess how well documents align with their assigned topics. The work targets public health communication data drawn from social media.
A cross-listed arXiv preprint describes a conservative, OCR-driven pipeline that reads food packaging images to check sodium and salt levels against South Africa's R214 limits. The authors note that screening decisions depend on establishing product identity, extracting nutrition facts panel evidence, choosing a reporting basis, and applying category-specific thresholds. The work frames automated label reading as a practical aid for regulatory compliance checks rather than a fully autonomous decision tool.
A new arXiv paper explores using generative diffusion models to forecast influenza incidence across space and time. The authors argue that existing mechanistic and statistical methods often fail to capture complex epidemic dynamics, and propose a generative approach as an alternative. The work targets public health planning, where better short-term forecasts can inform preparedness decisions.