Research And Application Of Microblog Text Sentiment Analysis Method Based On Deep Learning

As one of the main social platforms in China,more and more people are beginning to use microblog for information sharing and exchange.The microblog platform contains a lot of subjective texts,and timely sentiment analysis of microblog texts is of great significance for the government to understand the trend of public opinion and the product research of merchants.microblog texts have various forms,such as colloquialism and irregular grammar,which poses certain challenges to the sentiment analysis of microblog texts.Traditional deep learning methods have some problems in text sentiment analysis,such as insufficient semantic extraction and inability to highlight the role of keywords.In response to these problems,the main work of this paper is as follows:1.Aiming at the problem that bidirectional long short-term memory networks cannot solve the problem of polysemy and ignoring keywords,this paper proposes a microblog text sentiment analysis method of BERT-ABi LSTM.Firstly,the BERT pretrained language model is used to obtain the dynamic word vector representation of microblog text,then the Bi LSTM model is used to extract the semantic features of the text,the attention mechanism is introduced to calculate the importance of different semantic features.The experimental results show that compared with the Bi LSTM model,the accuracy rate has increased by 1.7%.2.Based on the BERT-ABi LSTM microblog text sentiment analysis model,in order to solve the problem that the model ignores the local features of text,this paper proposes a BERT-CNN-ABi LSTM microblog text sentiment analysis method.Firstly,the BERT model is used to obtain the word vector of the text,and then the improved Text CNN and ABi LSTM are used to extract the local features and global features of the text,and the local features are fused with the global features.The experimental results show that compared with the BERT-ABi LSTM model,the accuracy rate has increased by 0.6%.3.Based on the BERT-CNN-ABi LSTM microblog text sentiment model,design and implement the microblog text sentiment analysis system.Firstly,the demand analysis of the microblog text sentiment system is carried out,and then the overall design of the system is carried out according to the results of the demand analysis,and then the detailed design and implementation of the system are carried out,and finally the system is tested.