Jitendra Bhaware, Vivek Sharma
Fake reviews on online platforms pose significant challenges to consumers and businesses, undermining trust and decision-making. This paper introduces a two-phase hybrid model for fake review detection, leveraging the complementary strengths of Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks. In the first phase, CNN is employed to extract n-gram-like spatial features from text embeddings, capturing local patterns indicative of deceptive content. The second phase uses LSTM to model the temporal dependencies and sequential relationships within the reviews, enabling a deeper understanding of context and writing style. The hybrid architecture is trained on a labeled dataset of reviews, using pre-trained word embeddings to enhance feature representation and ensure robustness. Evaluation metrics such as accuracy, precision, recall, and F1-score demonstrate the model's superior performance over traditional machine learning and single deep learning approaches. This study highlights the effectiveness of integrating spatial and sequential feature learning for identifying fake reviews, offering a scalable and reliable solution for combating online deception.
Fake Review, Convolutional Neural Networks, Long Short-Term Memory, Accuracy, Precision, Recall, and F1-Score.