Machine Learning Based Sentiment Analysis of Movie Reviews——A case study of “Beyond the Clouds”

With the rapid development of Internet technology, major movie information platforms contain a huge amount of online review text data, mining and analyzing the viewpoints of the reviews as well as the hidden emotions have important commercial value. In this paper, we take the online theater reviews as the research object, obtain the review data of "Beyond the Clouds" through crawler, and use word cloud and word frequency analysis, positive-negative sentiment analysis, semantic network analysis and other methods to explore the sentiment information contained therein. The results show that moviegoers highly identify and recognize the real people and real deeds portrayed in such movies, and the content of user comments is basically dominated by positive emotions. Therefore, it is suggested that producers should pay attention to the relationship between the emotional tendency of film review data and film dissemination, and produce more films that reflect the mainstream ideology or that are close to the lives of ordinary viewers.