Detection of Fake Hotel Reviews Using ANFIS and Natural Language Processing Techniques

The wide spread of fake reviews presents a considerable threat to the correctness of online review systems which tends to undermining both business reputations and consumer trust. This research introduces an innovative method for identifying fraud hotel reviews through the use of an Adaptive Neuro-Fuzzy Inference System integrated with supervised learning algorithms. By integrating ANFIS with Random Forest, Support Vector Machines and Logistic Regression classifiers, we trained our model on a well-annotated dataset of hotel reviews. Our empirical analysis demonstrates that the ANFIS-based models significantly outperform traditional classifiers in terms of accuracy, recall, precision, F1 score and ROC-AUC metrics. It is observed that the ANFIS-SVM model achieved 92.5% accuracy, 91.8% precision. It also gives 93.2% recall, 92.5% F1 score and an ROC-AUC value of 0.95. These findings underscore the superior performance of our approach in categorizing authentic reviews from fake ones. Also the results show the impact of ANFIS-based models to improve the reliability of online review platforms, ultimately benefiting both businesses and consumers by enhancing the trustworthiness of user-generated content.