Effect of Aspect-Level Emotion Expression of Online Restaurant Reviews on Perceived Helpfulness: An Aspect-Based Sentiment Analysis Perspective

Few studies have explored how emotional expressions regarding various aspects affect review helpfulness despite having emotionally charged online reviews of various topics or aspects. The ultimate purpose of this study is to provide a comprehensive understanding of how different emotional expressions in online reviews influence their perceived helpfulness, specifically focusing on the moderating effect of restaurant type. To achieve this, we developed and empirically verified a research model using 85,117 online reviews collected from Yelp.com based on cue utilization theory. Specifically, it investigates the moderating effect of restaurant type on the relationship between emotional expression and review helpfulness in five aspects of online restaurant reviews (location, food, price, service, and ambiance). A BERT-based ABSA was performed to measure the emotional expression of each aspect of online restaurant reviews accurately, and a negative binomial regression model was used for the regression analysis. Emotional expressions regarding food and ambiance affected review helpfulness negatively. By contrast, services, prices, and locations affected review helpfulness positively. Negative reviews of food and ambiance were more helpful in luxury restaurants than in casual dining restaurants. In addition, a positive price evaluation had a more significant effect on review helpfulness for luxury restaurants than for casual-dining restaurants. Positive evaluations of service and location aspects significantly influence review helpfulness more in luxury restaurants than in casual-dining restaurants. Therefore, this study’s results can help restaurant practitioners to identify useful restaurant attributes and reduce information overload when potential consumers search for information.