Skip to main content
NUTRITION 2023 Main banner
Icon Legend
This session is not in your schedule.
This session is in your schedule. Click again to remove it.
Back

Topical Area: Nutritional Epidemiology

Nutritional Epidemiology (I) (Poster Session)

(P19-018-23) Social Media Mining for Food Safety Surveillance in Italy

Sunday, July 23, 2023
12:45 PM – 1:45 PM ET
Location: Poster Board 115

    Presenting Author(s)

  • Dario Gregori, PhD (he/him/his)

    Full Professor of biostatistics
    University of Padova
    Padova, Veneto, Italy

    • Co-Author(s)

  • SS

    Silvano Salaris

    Unit of Biostatistics, Epidemiology and Public Health (University of Padova)

  • GL

    Giulia Lorenzoni

    Unit of Biostatistics, Epidemiology and Public Health (University of Padova)

  • CL

    Corrado Lanera

    Unit of Biostatistics, Epidemiology and Public Health (University of Padova)

Disclosure(s):
Dario Gregori, PhD: No relevant financial relationship(s) with ineligible companies to disclose.


Objectives: Foodborne illness is a significant public health concern, with an estimated 600 million cases and 420.000 deaths each year worldwide. This preliminary work aims to prove that monitoring social data (Twitter) can be an alternative and complementary approach to the traditional methods to detect and prevent foodborne hazard.

Methods: Twitter data were analyzed using R as a coding language. Parameters were set to retrieve tweets in the last six years, written in Italian, with keywords related to foodborne pathogens. The frequency distribution of the tweets belonging to each pathogen keyword was generated. Data included in each peak above the calculated upper threshold value underwent a text mining procedure, by preprocessing them, cleaning the text from characters with no analytical value then generating a word cloud.

Results: The word cloud calculated from the tweets included in each peak returned specific information such as geographic areas, vehicle of infection, and food industries, allowing to link the data to an alleged contamination or foodborne outbreak. Seventy-six word clouds were clearly associated to well documented cases of outbreaks or food contamination, 55 word clouds returned confused or not related information to foodborne diseases, and in 2 cases possible events were not found through an online research.

Conclusions: Monitoring social data can be an alternative approach to detecting and preventing foodborne illness.

Funding Sources: PON Italia