From 5-Stars to Real Insight: Geospatial Detection of Campaigned Reviews with Google Maps and Mobility Data

In recent years, online review platforms have become increasingly popular, and many consumers rely on these platforms to search for restaurants before dining. However, some restaurants start long-term or short-term campaigns to boost their ratings and attract more customers to dine there. This behavior affects consumer decision-making and seriously undermines the fairness of the consumption market. Most existing research has primarily focused on the direction of reviews. In this study, we take the restaurant's perspective and propose a method to detect whether restaurants host campaigns. We introduce a series of features based on the overall restaurant information. To enhance the performance of our detection, we employ five different models and conduct a comprehensive evaluation. The experimental results from Google review data suggest that our method performs well in identifying restaurant campaigns.