Web text public emotion mining method and system based on supervised learning

The invention discloses a web text public emotion mining method and system based on supervised learning, and the method comprises the steps: building a coding framework of web text multi-dimensional public emotions according to broadcast study literatures, expert opinions and experiences; coding training is carried out on each bit of coding person, so that each bit of coding person is consistent; the web text is captured, and each coder performs scoring processing according to the given coding framework; extracting text features for expressing topics and contents in the web text by adopting an LIWC dictionary, and constructing a text data set; a supervised machine learning method is adopted, network text scoring and a text data set are combined for training, and a text analysis model is obtained through matrix parameter adjustment and cross validation methods; and obtaining evaluation subject attributes and multi-dimensional public emotion scores in the new web text by using the text analysis model. According