A perspective on data sharing in digital food safety systems.
Crit Rev Food Sci Nutr
; : 1-17, 2022 Jul 26.
Article
en En
| MEDLINE
| ID: mdl-35880485
ABSTRACT
In this age of data, digital tools are widely promoted as having tremendous potential for enhancing food safety. However, the potential of these digital tools depends on the availability and quality of data, and a number of obstacles need to be overcome to achieve the goal of digitally enabled "smarter food safety" approaches. One key obstacle is that participants in the food system and in food safety often lack the willingness to share data, due to fears of data abuse, bad publicity, liability, and the need to keep certain data (e.g., human illness data) confidential. As these multifaceted concerns lead to tension between data utility and privacy, the solutions to these challenges need to be multifaceted. This review outlines the data needs in digital food safety systems, exemplified in different data categories and model types, and key concerns associated with sharing of food safety data, including confidentiality and privacy of shared data. To address the data privacy issue a combination of innovative strategies to protect privacy as well as legal protection against data abuse need to be pursued. Existing solutions for maximizing data utility, while not compromising data privacy, are discussed, most notably differential privacy and federated learning.
Texto completo:
1
Colección:
01-internacional
Banco de datos:
MEDLINE
Tipo de estudio:
Prognostic_studies
Idioma:
En
Revista:
Crit Rev Food Sci Nutr
Asunto de la revista:
CIENCIAS DA NUTRICAO
Año:
2022
Tipo del documento:
Article
País de afiliación:
Estados Unidos