The comments on automobile purchasing platforms can effectively dig out consumers'views on new energy vehicles.Also,it can guide the market consumption to achieve a good transition from traditional vehicles to new energy vehicles.In this pa-per,a sentiment classification method for comment text,BS2CLA,is proposed.The sentiment word vector converted by Bert is weighted by sentiment dictionary,and the two-channel CNN model is used in combination with Bilstm-attention.Real comments of autohome are climbed and experiments are carried out on this basis.The results show that the model in this paper is better than the baseline model in F1,Accuracy and AUC.This model can better identify the emotion of a new energy vehicle and promote better and faster development of the market.
关键词
情感分析/新能源汽车/情感词典/双碳目标/深度学习
Key words
sentiment analysis/new energy vehicles/emotion dictionary/double carbon goal/deep learning
维普
万方数据