Strategic Insights from Customer Feedback: Study of Hotel Reviews Using Logistic Regression

This paper describes a systematic approach for analyzing a dataset of 20,490 hotel reviews collected from TripAdvisor.com using SAS. The goal of this study is to answer several business questions related to customer satisfaction in the travel industry, including identifying the top 10 most frequently used words in hotel reviews, the most common words across all reviews, and the most commonly referenced entities in the reviews. The insights of the analysis can help managers improve their offerings to meet their target visitors’ needs and desires, thus leading to improved customer satisfaction and retention. Additionally, the study incorporates Logistic Regression (LR) as a powerful machine learning algorithm to predict sentiments in hotel reviews. The LR model shows favorable performance. The obtained results showcase favorable LR model performance, indicating the model’s proficiency in distinguishing between positive and negative sentiments and correctly classifying the samples.