Foodborne Event Detection Based on Social Media Mining: A Systematic Review

Area of Science:

• Public Health
• Computational Epidemiology
• Data Science

Background:

• Foodborne illnesses pose a significant global health burden, with conventional surveillance methods often lacking timeliness.
• Early detection of foodborne disease outbreaks is crucial for effective public health interventions.
• Innovative approaches are needed to supplement traditional surveillance systems.

Purpose of the Study:

• To systematically review the role of social media platforms in detecting and managing foodborne illnesses.
• To evaluate the application of machine learning (ML) techniques in analyzing social media data for foodborne disease surveillance.
• To identify trends, challenges, and future directions in using social networks for public health monitoring.

Main Methods:

• Systematic literature search across major scientific databases (PubMed, EMBASE, Scopus, etc.) up to December 2024.
• Inclusion of studies utilizing social media data and data mining for foodborne disease prediction and prevention.
• Screening by independent reviewers, with data extraction on social media platforms, ML techniques (shallow and deep learning), and risk of bias assessment.

Main Results:

• Twitter and Yelp emerged as prominent data sources for foodborne illness surveillance via social media.
• Shallow learning models were predominantly employed in the reviewed studies.
• A significant proportion of identified studies exhibited a high or unclear risk of bias, indicating methodological limitations.

Conclusions:

• Social media combined with ML offers a valuable tool for real-time foodborne disease surveillance and outbreak response.
• There is a critical need for standardized methodologies to improve the reliability and validity of these approaches.
• Further research should focus on exploring the potential of deep learning models for enhanced predictive capabilities.