This paper proposes a novel deep learning (DL) model for detecting inconsistencies between customer reviews and ratings on e-commerce websites. The model uses a two-branch DL architecture to classify customer reviews as positive or negative and assess the consistency between these classifications and actual ratings. The model is evaluated on a large dataset of Amazon product reviews, demonstrating superior performance in prediction accuracy and other metrics compared to existing models. The model's effectiveness is validated through extensive experiments, showing that it can identify inconsistencies between reviews and ratings, helping customers make more informed purchasing decisions. The proposed DL model combines features such as increased representation power, improved learning efficiency, and flexibility in branch design. It uses two parallel branches to process input data, enhancing the model's ability to learn complex patterns and relationships. The model's performance is evaluated using standard metrics including accuracy, precision, recall, F1-score, sensitivity, and specificity. The results show that the proposed model outperforms existing methods in terms of classification accuracy and other performance measures. The model is also used to examine the consistency between customer reviews and their actual ratings, identifying discrepancies that may indicate unreliable reviews. The study highlights the importance of detecting inconsistencies in customer reviews and ratings to improve the reliability of e-commerce platforms. The proposed model has potential applications beyond e-commerce, including fake news detection, natural language processing, and fraud detection. The model's ability to handle large volumes of data and its efficiency in processing complex information make it a valuable tool for various domains. The study also identifies limitations of the model, including the need for further validation on additional platforms and the challenge of accurately addressing nuanced language issues such as sarcasm and irony. Future research directions include exploring the model's performance in other languages and incorporating more advanced machine learning techniques for further analysis. The proposed DL model provides a robust solution for detecting inconsistencies in customer reviews and ratings, enhancing the reliability of e-commerce platforms and improving the customer experience.This paper proposes a novel deep learning (DL) model for detecting inconsistencies between customer reviews and ratings on e-commerce websites. The model uses a two-branch DL architecture to classify customer reviews as positive or negative and assess the consistency between these classifications and actual ratings. The model is evaluated on a large dataset of Amazon product reviews, demonstrating superior performance in prediction accuracy and other metrics compared to existing models. The model's effectiveness is validated through extensive experiments, showing that it can identify inconsistencies between reviews and ratings, helping customers make more informed purchasing decisions. The proposed DL model combines features such as increased representation power, improved learning efficiency, and flexibility in branch design. It uses two parallel branches to process input data, enhancing the model's ability to learn complex patterns and relationships. The model's performance is evaluated using standard metrics including accuracy, precision, recall, F1-score, sensitivity, and specificity. The results show that the proposed model outperforms existing methods in terms of classification accuracy and other performance measures. The model is also used to examine the consistency between customer reviews and their actual ratings, identifying discrepancies that may indicate unreliable reviews. The study highlights the importance of detecting inconsistencies in customer reviews and ratings to improve the reliability of e-commerce platforms. The proposed model has potential applications beyond e-commerce, including fake news detection, natural language processing, and fraud detection. The model's ability to handle large volumes of data and its efficiency in processing complex information make it a valuable tool for various domains. The study also identifies limitations of the model, including the need for further validation on additional platforms and the challenge of accurately addressing nuanced language issues such as sarcasm and irony. Future research directions include exploring the model's performance in other languages and incorporating more advanced machine learning techniques for further analysis. The proposed DL model provides a robust solution for detecting inconsistencies in customer reviews and ratings, enhancing the reliability of e-commerce platforms and improving the customer experience.