Sentiment analysis based on CNN - LSTM hotel reviews

Abstract In order to obtain the implied semantic information of hotel reviews for emotional analysis, correlation between discontinuous words is ignored in traditional convolutional neural network (CNN) analysis. Therefore, a novel sentiment analysis method based on CNN - LSTM model proposed. this method, used extract features from review texts, and add sentence structure enhance deep learning. This improves accuracy F1 value CHN senticorp-HTOL-BA-6000 data set, can better solve task text discover orientation information.