Why is the human element important when working with food data?

We have over 100.000 products in our system. Since we’re working with food it’s highly critical that all the information is correct and that we have good-quality data. Our algorithm interprets and structures each product down to the ingredient level, and our platform can then compare ingredients and enrich the product data with other sources or automate calculations. This automation speeds up the process, I’d say it does 90% of the work. But that last 10% needs a human eye to ensure the data quality.

That’s where our data quality team comes in. After a product has run through the algorithm it’s given a score. If the score is 1 the program recognizes all the ingredients and has understood the descriptive words. If it is below 1, our data quality team checks the product. Sometimes our algorithm is unable to identify an ingredient because of a typo or is confused by various messages within the ingredients such as “with a splash of lemon juice” or a uniquely phrased allergen statement. Not to forget all the new ingredients that keep popping up these days! In these cases, our data team will make the necessary corrections and add ingredients that potentially were missing to make sure it all adds up. We also make sure that all the information is tagged correctly so that our clients can fully rely on the product data.

Tell us about your journey and you ended up at FoodFacts?

I've always been interested in the human body and health. In my teens, I wanted to become a nutritionist, which then shifted into a passion for working with health through movement and the path of physical therapy. Nevertheless, my interest in food and what it contains has remained. Throughout the years I've been an avid seeker of knowledge, studying the food labels in grocery stores closely. When I found FoodFacts, I knew it would be a perfect fit. I could continue learning and working with food data professionally. And at the same time empowering others to discover and spread this valuable information. Since I enjoy working with people, managing the data quality team brings it all together wonderfully. 

How can food data enable better food choices and transparency?

Even though there are regulations around how ingredient lists should be written, there are still many misunderstood ingredients which often leads to the information on food packaging being confusing or even misleading. 

Ingredients with negative connotations can be hidden, for example, adding more sugar to a product by naming it "starch syrup" or calling African Palm Oil “Elaeis Guineensis”. Far from all consumers have the time or energy to keep track of all of these ingredients and how they might affect their health or the planet. This is where FoodFacts come in and help the consumer to assess a product quickly, by identifying and highlighting certain aspects, both related to health or sustainability. Our automated tag “added sugar” or “palm oil free” would detect this and could be used to improve a digital search filter or automated product swap

What food fact do you think everyone should know?

No food group is inherently evil or bad for you. Sure, we all have foods that work better or worse for us individually, and it's always about balance. But no food group deserves to be demonized. What's important is our relationship to food, how we relate to it and experience it, and being able to enjoy the soul-nourishing moments with all those foods that make us feel good.