Artificial Intelligence and Amazon Reviews Could Help Detect Unsafe Food

Researchers have trained machine learning algorithms to sniff out unsafe food products that would be recalled by the US Food and Drug Administration (FDA).

13.08.2019 | by Kezia Parkins
Photo by Franki Chamaki on Unsplash
Photo by Franki Chamaki on Unsplash

Annually, the Food and Drug Administration (FDA) has to recall hundreds of food products, such as cookies for kids containing chunks of blue plastic, salmonella-tainted spinach and baby formula with a “presence of metal foreign material.”

The Centers for Disease Control and Prevention estimates that 76 million foodborne illnesses, including 325,000 hospitalizations and 5,000 deaths, occur each year in the United States. According to the United States Department of Agriculture, foodborne illness costs the US economy $10 to 83 billion per year.

Access to safe and nutritious food is essential for good health.

However, food can become unsafe due to contamination with pathogens, chemicals or toxins, or the mislabeling of allergens. Illness resulting from the consumption of unsafe foods is a global health problem. 

It can sometimes take months before a contaminated or unsafe product is recalled by the FDA. 

But now, researchers have come up with a method that might fast-track that process, leading to early detection and, ultimately, faster recalls. 

More than ever, we are buying our food online, and online shoppers tend to leave reviews. These act like breadcrumbs for food safety officials looking for hints to find unsafe food products.

Researchers have developed a machine learning approach for detecting reports of unsafe food products in consumer product reviews from Amazon.

They linked Amazon food reviews to FDA food recalls from 2012 to 2014 using text matching approaches in a Postgres relational database.

They then trained machine learning algorithms to differentiate between reviews for recalled items and reviews for items that had not yet been flagged. 

The results of the study showed that the algorithms were able to predict FDA recalls three-quarters of the time and also identified another 20,000 reviews for possibly unsafe foods— suggesting that there may be many more products should have been recalled or investigated.

The results are published in the Journal of the American Medical Informatics Association. [Adyasha Maharana, Detecting reports of unsafe foods in consumer product reviews]

Join the discussion

Leave a Reply

Related Shakers

Oluseun Onigbinde


Robert Scott Lazar

United Nuclear

Related Shakers