Text classification method, electronic device and computer storage medium

Abstract

The present invention provides a text classification method, electronic equipment and computer storage medium, wherein, the method includes: acquiring multiple topic categories and corresponding multiple words and multiple documents; counting all documents of each topic category containing The number of first documents of each word corresponding to the topic category and the number of second documents containing each word of each topic category in all documents of all topic categories; calculate the number of first documents of each word of each topic category The ratio of the number of documents to the second document is used as the word-to-topic coverage; if the word-to-topic coverage of the word is greater than the set threshold, the word is selected as the feature word of the corresponding topic category, and the corresponding category feature word bag is obtained; the classification document is treated Word segmentation to obtain the bag-of-words model of the document to be classified; calculate the similarity between the bag-of-words model of the document to be classified and the bag of feature words of each category; determine the category of the document to be classified according to each similarity. Through the above scheme, the text classification task can be completed with fewer annotations.

Description

Technical field [0001] The present invention relates to the field of natural language processing, and more particularly to a text classification method, an electronic device and a computer storage medium. Background technique [0002] Text classification is a traditional natural language processing problem, which has a large value of practical value, is a supporting technology that realizes spam filtering, news automatic archiving, text data mining and other tasks. The classic text classification problem is the process of imparting reasonable category labels for new text after learning from high quality labeling. Traditional text classification method focuses on the statistical characteristics of the text, the shallow semantic characteristics, such as TF-IDF, N-GRAM model, word embedding, etc., and establish text feature vectors, combined with classification algorithms to implement text classification tasks, classification algorithms include simple Ye Lus, logic regression, suppo...

AI Technical Summary

Problems solved by technology

[0006] These are mainly used in scenarios that lack annotation, require a lot of manual feature engineering costs, rely on external knowledge bases, expert knowledge, and synonyms, etc., and have a relatively low level of automation and intelligence. Poor performance on multiple evaluation indicators such as

Images