Improved deep learning model text classification

Deep learning technology develops rapidly. Convolutional Neural Network (CNN), as a key technology in deep learning, has been favored and concerned by many scholars and widely applied in information retrieval, classification, data management, mining and other fields. In order to fully obtain the local features and key words of text, a TCNN-DAM model is proposed based on the study of TCNN model, which aims at maximizing the representation of text features, improving the text classification effect, and promoting the model to better classify in sogou news corpus. Tests show that the improved model has outstanding classification effects, which can effectively improve the accuracy, precision, recall and F
₁
value of classification.