Recent Papers

Online customer reviews have become an increasingly influential tool in shaping purchasing decisions. However, the growing impact of these reviews has led to a surge in the publication and promotion of fake reviews by some businesses, either to enhance their own product's reputation or to undermine their competitors. These counterfeit reviews can have an especially detrimental impact on small businesses, with even a single negative fake review capable of causing significant damage. In this context, the current study introduces a technique for classifying and identifying fake reviews using machine learning (ML) methodologies. The proposed algorithm was applied to the Yelp dataset for hotel services. The text was initially preprocessed through four stages: tokenization, normalization, stop word removal, and stemming. Subsequently, features were extracted using TFIDF techniques to leverage the benefits of sentiment analysis and to ascertain the presence of spam comments in the feature extraction approach. During the classification phase, the study employed three ML algorithms: Xgboost, a support vector classifier, and stochastic gradient descent. The proposed model was evaluated on both balanced and imbalanced datasets, using oversampling and undersampling techniques to determine its accuracy. The findings of this research hold promise for enhancing the credibility of online reviews and protecting businesses from the adverse effects of fake reviews. By unmasking fraudulent reviews, this study contributes to ensuring the integrity of online review platforms and safeguarding the interests of both businesses and consumers.

Copyright © 2024 Mrs. Jayashri Kunde. This is an open access article distributed under the Creative Commons Attribution License.