Google has developed a new algorithm in collaboration with researchers from Harvard University that can flag restaurants that could possibly leave you with food poisoning. The algorithm is capable of flagging lapses in food safety in almost real time. It has tested this algorithm in Las Vegas and Chicago to cross reference users’ food poisoning-related search terms with saved location history data from smartphones about restaurants recently visited by the user.
Some restaurants that were flagged by Google’s algorithm as potentially unsafe were then visited by health inspectors. They also visited restaurants that were flagged by more conventional methods such as complaints from patrons. The inspectors didn’t know which restaurant was flagged by which method.
They then compared the algorithm’s results with routine inspections conducted by health departments in the two cities. The rate of potentially unsafe restaurants flagged by the algorithm was 52.3 percent compared to the overall rate of detection by routine inspections which was 22.7 percent across the two cities.
“In this study, we have just scratched the surface of what is possible in the realm of machine-learned epidemiology,” said Google research scientist and co-author of the study Evgeniy Gabrilovich. The algorithm could certainly play a major role in the fight against foodborne illness in the future.
Filed in . Read more about Google and Machine Learning. Source: nature