Research Progress of Sentiment Analysis Based on Deep Learning

Sentiment analysis technology is widely used in film and television media, social media, entertainment industry, medical and other fields, and has great commercial use value. In the early stages of research, sentiment analysis methods based on traditional machine learning are not effective, which greatly limits its application in the commercial field. In recent years, with the development of artificial intelligence, deep learning-based sentiment analysis have made great advances. However, the large variation in performance results and characteristics for different algorithms heightened the difficulty in practical application. According to the technical progress of sentiment analysis, this paper summarizes from two aspects: data set and algorithm, introducing the commonly used data sets, evaluation indicators and sentiment analysis methods based on deep learning. In addition, this paper also provided a comparison among different algorithms widely used for sentiment analysis. Finally, this paper analyzes the existing technical adjustment, and discusses the future research direction.