This paper proposes an Internet public sentiment analysis system based on data fusion. The news
We Chat public information
blog
forum
APP
micro-blog
newspapers
video which collected from the Internet by web crawler and DPI which from the China Mobile had been used to achieve data fusion and data analysis by using multiple Natural Language Processing algorithms such as the sentiment analysis algorithm. Multi-tenant had been used to achieve the data sharing and information isolation
which can improve resource utilization and allow user to experience personalized public opinion.