Automatic text sentiment analysis method based on AM-CNN algorithm

The invention discloses an automatic text sentiment analysis method based on an AM-CNN algorithm, wherein the method comprises the following steps: 1, crawling curriculum evaluation information, storing the curriculum evaluation information in a MySQL database, and exporting the data into a csv file; 2, reading csv file data by using a pandas library function, and carrying out preprocessing and word vector conversion; 3, dividing the converted word vectors into a training set I and a test set, and generating training sets II from the training set I through attention mechanisms; 4, respectively sending various training sets II into a convolutional neural network, carrying out feature extraction on data under each attention mechanism, and extracting important information of a data matrix by a pooling layer; and 5, integrating the pooled data by a merging layer, and sending the pooled data to a full connection layer and an output layer for processing. According to the method, the attention mechanisms are used fo