To continuously improve the quality of the data we rely on 4 pillars:
1. The community is constantly adding data and photos and is a veritable patrol of reviewers.
2. We have set logic rules to identify errors on the product sheets. For example: if the sum of the weight of the ingredients is greater than the total weight of the product, there is an anomaly. We have around 50 others quality checks.
3. Thanks to Machine Learning, we limit the risk of input errors. Our technology allows us to extract textual data from photos taken by contributors
4. Thanks to the support of Santé Publique France, we developed a portal allowing producers to upload their mass-produced data and thus correct/complete the community's contributions.