A smart recommendation model for airline reviews via deep learning sentiment analysis (text ratings fusion with LSTM)
TL;DR
The LSTM model achieved the highest accuracy of 91% and an area under the curve (AUC) of 90%, outperforming the other models and enabling more accurate sentiment-based predictions and enhanced recommendation quality.
Abstract
This study develops a smart recommendation model based on deep learning (DL) for sentiment analysis of airline customer feedback. It uses textual content and ratings to provide personalized recommendations. We evaluate four DL models— Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and 1D Convolutional Neural Network (COV1D) —for sentiment analysis of airline customer feedback. The LSTM model achieved the highest accuracy of 91% and an area under the curve (AUC) of 90%, outperforming the other models. The results show that all models achieved significant improvements in accuracy, precision, recall, and F1-score compared with baseline methods. By analyzing sentiment polarity in user reviews, the model inferred user preferences and the relationship between sentiment and ratings. Overall, its ability to capture textual context enabled more accurate sentiment-based predictions and enhanced recommendation quality.