Chick-fil-A's custom artificial intelligence-powered system uses customers' social media posts to identify food safety issues at its more than 2,400 restaurants across the U.S. with up to 78 percent accuracy, according to VentureBeat.
Davis Addy, the fast food chain's senior principal IT leader of food safety and product quality, presented the tool at the ReWork Deep Learning Summit in Boston on May 23. The algorithm continuously collects comments about the company from a variety of social platforms, filters them for any of 500-plus food safety-related keywords, then estimates their legitimacy by identifying underlying sentiments and connotations.
Chick-fil-A worked with Amazon's AWS Comprehend natural language processing system to refine the system. The algorithm is therefore able to detect sarcasm, slang and falsehoods, as well as misspelled variations of crucial words such as "filthy" and "illness," enabling it to identify food safety risks and predict outbreaks of foodborne illnesses with nearly 80 percent accuracy.
More articles about AI:
Cardinal Analytx raises $22M for health spending prediction tool
Viewpoint: HIPAA, a law from 1996, will limit AI in healthcare
Pitt, Carnegie Mellon creating robotic trauma care system for combat casualties