Analysis of the Effect of Online Reviews on Restaurant Rankings

This study analyzes the impact of consumers' opinions included in online reviews on restaurant rankings using CRT of the decision tree model among data mining methodologies. The rankings of restaurants located in Jeju Island were used as the dependent variable, and the number of reviews, ratings, quality evaluation, sentiment score of the reviews, foreigners' review ratio, and the operation periods of restaurants were included as the explanatory variables. Data such as reviews and ratings were collected from Tripadvisor through web crawling using Python, and the sentiment score of the reviews was computed using the VADER package. Results show that the variables that best explain restaurant rankings were the number of reviews and foreigners' quality evaluations. The higher these values were, the higher the restaurants were ranked. The academic significance of this study is that we collected the information containing online reviews and analyzed its impact on restaurant rankings. This study has the distinct features of adding the sentiment score in the model while inclusively using variables related to online reviews used in existing studies, and comparing the impact of the review characteristics of Koreans and foreigners on restaurant rankings. This study's results reveal that online platforms are the key in informing potential visitors and boosting restaurant visibility, emphasizing the need for review-focused marketing. The number of reviews significantly affects consumer visiting decisions, suggesting restaurants should strategize to garner more reviews. Furthermore, the quality of evaluations made by foreign tourists impacts restaurant rankings more than those of local tourists'. This underscores the importance of services and marketing specifically aimed at foreigners, especially in regions frequented by them. The implementation of these strategies could attract more tourists and increase competitiveness in these areas.