Localized Authenticity and Review Credibility in Japanese Restaurant Reviews: A Computational Analysis of Dianping Data from Beijing and Shanghai

This study examines how diners evaluate Japanese restaurants on Dianping, a major Chinese restaurant-review platform, focusing on how localized authenticity, expectation mismatch, and review credibility are reflected in review language. Localized authenticity refers to how diners judge whether a locally adapted Japanese restaurant remains culturally recognizable, while expectation mismatch captures gaps between platform-generated expectations and offline dining experience. Review credibility reflects instances where such mismatches develop into doubts about ratings, photos, or prior reviews. Using 230,185 reviews from Japanese restaurants in Beijing and Shanghai between 2020 and 2024, the study combines descriptive analysis, keyword-assisted thematic coding, and logistic regression. The analysis links review themes such as freshness, service, atmosphere, price, and queueing to both rating outcomes and credibility-related language. The results show systematic differences in how Japanese restaurants are evaluated across the two cities, with Beijing reviews placing more emphasis on environment and Japanese-style cues, and Shanghai reviews focusing more on price, service, and value. Negative reviews are more strongly associated with operational friction and mismatch language than with food-related cues alone. Expectation-gap language increases over time, while direct credibility-related language remains rare and concentrated in low-rated reviews. Overall, the findings suggest that Japanese restaurant reviews reflect a connected process in which cultural interpretation, service experience, and platform-generated expectations jointly shape consumer evaluation.