Sentiment Analysis and Authenticity Assessment of Hotel Reviews

Travelers frequently base their judgments on online hotel ratings; however, the presence of faux reviews seriously weakens the credibility of these sites. The major focus of this research is the development and testing of machine learning algorithms for detecting fake hotel reviews. The approaches employed include Multinomial Naive Bayes (MNB), Support Vector Machine (SVM), and Logistic Regression models, all of which have a good reputation for text categorization efficiency. We have used the dataset, which included 9479 reviews, both fake and legitimate. Text preprocessing and feature engineering were used, and the text was then translated into numeric vector using the TF-IDF vectorization and Count Vectorization approach. On the dataset, each method was trained and tested using typical assessment measures like accuracy, precision, recall, and F1-score. The results reveal that all three algorithms perform effectively in discriminating between actual and false reviews. However, the SVM model has a maximum accuracy of 82% and is thus one of the best suitable algorithms to identify fraudulent reviews in the hotel industry. This study will help customers and the hospitality industry to make informed decisions and lessens the impact of misleading reviews.