Many products do not list the amount of fiber, which will result in an overestimate of the Nutrition score.
Proposed solution
We could use an estimate for the amount of fiber to get a better Nutrition score. This value can either be hard-coded into the category taxonomy, or based on the median value of the category.
Additional context
The issue is triggered by a clean up of the walnut category. Many products have a Nutrition score above 0, i.e. Nutriscore B or higher. These products seem to miss the fiber value. Subtracting 5 points for the usual fiber, would turn these products into Nutriscore A.
Number of products impacted
How many products do not have a fiber listed, but do have a non-negligable amount of fiber.
Problem
Many products do not list the amount of fiber, which will result in an overestimate of the Nutrition score.
Proposed solution
We could use an estimate for the amount of fiber to get a better Nutrition score. This value can either be hard-coded into the category taxonomy, or based on the median value of the category.
Additional context
The issue is triggered by a clean up of the walnut category. Many products have a Nutrition score above 0, i.e. Nutriscore B or higher. These products seem to miss the fiber value. Subtracting 5 points for the usual fiber, would turn these products into Nutriscore A.
Number of products impacted
How many products do not have a fiber listed, but do have a non-negligable amount of fiber.