A method and server for classifying text information

Abstract

The invention discloses a method for classifying text information, which comprises the following steps: obtaining the word bag information and the word vector information of the first text information; acquiring a topic distribution matrix corresponding to the word bag information via a neural network topic model, wherein the neural network topic model comprises an encoding module and a decoding module and the topic distribution matrix is obtained by encoding the bag information by the encoding module and decoding module in the process of decoding the bag information; generating a second textinformation according to the topic distribution matrix and the word vector information corresponding to the word bag information; processing the second text information by a classifier to obtain a text information classification result, wherein the classifier is trained according to the text information set to be trained. The invention also discloses a server. The invention can be widely used in the application field, and the potential subject distribution of the text information can be accurately learned by using the neural network subject model, the distribution can assist the classificationof the text information, and the credibility of the classification can be improved.

Description

technical field [0001] The invention relates to the field of artificial intelligence, in particular to a method and server for classifying text information. Background technique [0002] With the rapid development of the Internet, Internet entertainment has become a hobby of many people. In Internet business, it is very important to classify text information. Whether in the e-commerce industry, social platforms or search engines, the classification of text information can be effectively applied to tasks such as information retrieval, sentiment analysis, public opinion monitoring and event tracking. [0003] Currently, text information can be classified based on external data information. This type of method uses an external information data source, such as Wikipedia or knowledge base, to expand the features of the current text information, and output the text classification results of the text information through the neural network. [0004] However, although the method o...

AI Technical Summary

Problems solved by technology

[0004] However, although the method of generating classification results based on external data information can solve the problem of data sparseness, it is still limited in the application field, especially for text information like social media, which often has a fast update speed of text information and The characteristics of non-standard text information, its generalization ability will be limited by the source of information, thus reducing the credibility of classification

Images