About
A new research project, led by Elad Yom-Tov, a Computer Science Professor at Bar Ilan University, in collaboration with University College London (UCL) and a UK national public health institute, focuses on identifying hotspots of gastrointestinal diseases by analyzing the online reviews of vacationers. This innovative approach aims to create a rapid detection method for areas of increased risk for disease, enabling more timely and effective public health interventions.
Challenge
Traveling can enable the spread of pathogens, but traditional health surveillance systems are often blind to threats from outside their geography. Professor Yom-Tov and his research team are addressing the need for innovative, data-driven methods to detect health threats that may not be captured through traditional surveillance systems. Leveraging Bright Data’s Google Maps reviews dataset and Web Scraper API, the research covers approximately 3350 hotels across 20 popular vacation destinations and validates findings against known outbreak data to detect potential foodborne illness outbreaks associated with all-inclusive hotels in various countries.
Around 0.1% of hotels include a mention of food poisoning attributed to the hotel’s restaurants. However, some vacation destinations have 3 times as many mentions of food poisoning. At some of the worst hotels over 3% of their reviews discuss food poisoning, many of them leading to requiring medical care.
Impact
This project has the potential to significantly enhance public health response capabilities by enabling the early detection of foodborne illness outbreaks. Beneficiaries include public health agencies, local authorities, and travelers who may receive appropriate treatment recommendations. The project aims to produce a prototype and an analytical report that could influence policy, improve hotel sanitation practices, and contribute to scientific literature. By following a value-driven approach prioritizing actionable insights, public safety, and responsible digital data use, the project could expand its methodology to other regions and health conditions, demonstrating the broader utility of internet data in epidemiological research.
