Using social media to detect public health rumors

Abstract

The contemporary epidemiological scenario, with its rapid dissemination of information and
misinformation, necessitates enhancements in health surveillance systems, particularly in detecting rumors.

Platforms such as social media offer a large volume of data that, although challenging to access, can be uti-
lized to monitor health rumors. This study explores the application of technological resources and proposes

a conceptual framework to optimize the utilization of social media information by health surveillance teams,
facilitating the analysis and detection of rumors. Validation of the framework, using practical examples
with Twitter data, identified a significant set of messages relevant to the detection of rumors related to
diarrheal and respiratory syndromes, thereby optimizing analysis time compared to manual searches. We
conclude that the proposed framework offers a structured and promising approach to optimizing the use of
social media information in health surveillance, aiding in the rapid identification of rumors. This research
contributes to the development of a mechanism for classifying these messages, the formulation of adaptable
search and analysis plans, and the creation of repositories for managing and reusing data and parameters,
thereby increasing the efficiency of public health surveillance.