Text classification algorithm based on a circulating neural network variant and a convolution neural network

Abstract

The invention discloses a text classification algorithm based on a circulating neural network variant and a convolution neural network, The algorithm includes step 1, preprocessing that data set SogouC and the data set THUCNews, dividing the two data set into a training set and a test set, and training the text data in the respective training set and the test set into sentence vectors; step 2, using the training set text of the two datasets in Step 1 to establish the BGRU-CNN hybrid model; step 3, establishing an objective function, and using a random gradient descent method to train that BGRUestablish in the step 2; step 4, inputting the text sentence vector of the test set into the BGRU trained in step 3 in the two data sets. The classification results are obtained. The invention solvesthe problems that the difficulty of extracting the semantic key features is great and the classification effect is poor when the long text is classified in the prior art.

Description

Technical field [0001] The invention belongs to the technical field of natural language processing methods, and relates to a text classification algorithm based on a cyclic neural network variant and a convolutional neural network. Background technique [0002] At present, the Internet is developing rapidly, producing a large amount of text information every moment. How to effectively classify and manage a large amount of text, and then quickly understand the value of information, is the focus of many researchers. Among them, the long text contains many different keywords. Therefore, in text classification, keeping the structure of the long text intact, maintaining the order between text words, and learning the text context semantics can improve the text classification effect of the long text. [0003] Text classification mainly includes text representation, selection and training of classifiers, evaluation and feedback of classification results. The text representation is a key...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to provide a text classification algorithm based on the variant of the cyclic neural network and the convolutional neural network, and combine the variant of the cyclic neural network with the convolutional neural network to solve the problem of long texts existing in the prior art in text classification. It is difficult to extract semantic key features at the same time, and the classification effect is poor

Images