Early detection of foodborne illnesses in social media

Casas, Jacky (School of Engineering and Architecture (HEIA-FR), HES-SO // University of Applied Sciences Western Switzerland) ; Mugellini, Elena (School of Engineering and Architecture (HEIA-FR), HES-SO // University of Applied Sciences Western Switzerland) ; Abou Khaled, Omar (School of Engineering and Architecture (HEIA-FR), HES-SO // University of Applied Sciences Western Switzerland)

Alert Center is a platform aiming at detecting outbreaks caused by food toxin infections and food intoxications in Switzerland. It does this by analyzing tweets and sending alerts to the Federal Food Safety and Veterinary Office (FSVO) when a risk is detected. The platform is composed of four main parts: a real-time extractor that targets tweets based on a list of curated keywords, three classifiers (one for each main spoken language) that isolate tweets related to food toxin, a system that locates tweets on the Swiss territory and a web-based dashboard to visualize the results. Combining localization algorithms of tweets and users allows the system to locate 75.09% of the tweets, 2.31% of which were located in Switerzland. In addition, a list of Swiss Twitter accounts corresponding to 15% of the total estimated number of Swiss accounts has been created.


Keywords:
Conference Type:
published full paper
Faculty:
Ingénierie et Architecture
School:
HEIA-FR
Institute:
HumanTech - Technology for Human Wellbeing Institute
Publisher:
Lausanne, Switzerland, 23-25 April 2020
Date:
2020-04
Lausanne, Switzerland
23-25 April 2020
Pagination:
6 p.
Published in:
Proceedings of the 2nd International Conference on Human Interaction and Emerging Technologies: Future Applications (IHIET – AI 2020), 23-25 April 2020, Lausanne, Switzerland ; Advances in Intelligent Systems and Computing
Numeration (vol. no.):
2020, vol 1152, pp. 415-420
DOI:
ISSN:
2194-5357
ISBN:
978-3-030-44266-8
Appears in Collection:



 Record created 2021-01-12, last modified 2021-01-20

Fulltext:
Download fulltext
PDF

Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)