Small sample multi-label text classification model training method and text classification method

Abstract

The invention discloses a small sample multi-label text classification model training method, a small sample multi-label text classification method and device, computing equipment and a computer storage medium. According to the method, the key phrases are extracted from the labeled sample text corresponding to the text label, the prompt template is expanded according to the key phrases, and the data augmentation is realized based on the prompt template, so that model training based on the labeled sample text of a small sample is realized, the defect that a large scale of labeled texts cannot be obtained is overcome, and the user experience is improved. Moreover, the training efficiency of the text classification model is improved, it is not needed to spend a long time for manual labeling, and the labor cost and the time cost are reduced.

Description

technical field [0001] The present invention relates to the field of computer technology, in particular to a training method for a small-sample multi-label text classification model, a small-sample multi-label text classification method, a device, a computing device and a computer storage medium. Background technique [0002] Text classification is a classic task in the field of natural language processing, which aims to automatically label text with one or more pre-defined class labels using machines. After the rise of deep learning technology, text classification research has made great progress. However, the high accuracy of deep learning relies on a large amount of labeled data, and data labeling requires a lot of labor and time costs. In many cases, large-scale labeled data is not available. When the number of text categories is large, the labeled data The number of texts belonging to each category will have a long-tailed distribution, and the number of texts in most c...

AI Technical Summary

Problems solved by technology

However, the high accuracy of deep learning depends on a large amount of labeled data, and data labeling requires a lot of labor and time costs. In many cases, large-scale labeled data is not available. When the number of text categories is large, the labeled data The number of texts belonging to each category will be distributed in a long tail, and the number of texts in most categories is very small. Therefore, there is an urgent need for a text classification scheme based on small-scale labeled texts

Images