Analysis of Emotional Influencing Factors of Online Travel Reviews Based on BiLSTM-CNN

This paper analyzes the influencing factors of tourism development through the emotional tendencies in online travel reviews, and uses a Bidirectional long short-term memory Convolutional Neural Network (BiLSTM-CNN) deep learning model to classify online travel reviews. The model has high accuracy and good loss function convergence. Then we use the Dynamic Topic Models (DTM) model to analyze the classified texts at two levels. At the micro level, the main influencing factors of a destination are obtained for a certain destination, and corresponding improvement plans are proposed for the negative influencing factors. At the macro level, this paper analyzes the changing trend of the destination’s emotional inclination under the two influencing factors of fare and traffic.