Assessing the Importance of Key Restaurant Review Topics Using TF-IDF and Entropy Method

Customer reviews play a crucial role in shaping a company’s reputation, informing decision-making processes, and influencing overall business strategies. While sentiment analysis and aspect-based sentiment analysis are widely used in the literature to classify comments as positive, negative, or neutral, limited research has focused on extracting key topics from reviews and determining their relative importance. This study aims to address this gap. We identify four main topics—service quality, price, ambience, and food/taste—based on the most significant words determined through Term Frequency-Inverse Document Frequency (TF-IDF) scores. After assigning high TF-IDF score words to their corresponding topics, the entropy method, a commonly applied technique in multi-criteria decision-making, is utilized to evaluate the importance of each topic based on customer feedback. The relative importance of topics is derived by aggregating TF-IDF scores within each comment. The findings reveal that price has the highest weight, while food/taste has the lowest. These results suggest that price-related comments are less prevalent compared to food/taste-related discussions. Furthermore, the findings highlight the need for greater attention to be given to service quality and ambience, as they hold significant importance in customer perceptions.