Text classification method and device, terminal and storage medium

Abstract

The invention provides a text classification method and apparatus, a terminal and a storage medium. The method comprises the steps: obtaining text data; performing word segmentation processing on thetext data through a preset Bert text processing model to obtain word vectors; extracting context feature data corresponding to each word vector from the text data through a preset context feature extraction model; and inputting the context feature data into a preset capsule network classification model for topic classification operation, and outputting a classification result. According to the application, a method of combining Bert, GAT and a capsule network is adopted; pretreatment by Bert, the GAT learns text context features; carrying out theme classification on the features through a capsule network; carrying out a dynamic routing process in the capsule network; according to the method, some defects possibly brought to the text classifier model in the back propagation process of datain the convolutional neural network are reduced, so that the text classification accuracy is improved, and the technical problem that an existing deep learning method is low in text classification accuracy is solved.

Description

technical field [0001] The present application relates to the field of text classification, in particular to a text classification method, device, terminal and storage medium. Background technique [0002] With the development of the big data era, a large amount of text data has been accumulated, and the text data involves a variety of categories and topics. In the general field, through the analysis and processing of these text data, we can gain a deeper understanding of the relationship among them. However, the various types of texts obtained are diverse, difficult to be understood by computers, and involve different themes. Assuming that these texts can be effectively classified and identified, it will have a great positive effect on big data processing. [0003] At present, the mainstream methods for text classification are: GNN, CNN. Although the cyclic neural network GNN can obtain the contextual features of the text, it has the problem of gradient explosion; the conv...

AI Technical Summary

Problems solved by technology

[0004] This application provides a text classification method, device, terminal and storage medium, which are used to solve the technical problem of low text classification accuracy in existing deep learning methods

Images