Cross-field text classification method based on self-adaptive noise reduction encoder

Abstract

The invention discloses a cross-field text classification method based on a self-adaptive noise reduction encoder. The method includes: using a feature selection method suitable for use in a cross-field task to filter out feature words with low appearing frequency and meaninglessness in samples in a source-field data set and a target-field data set, calculating a better noise interference coefficient in a self-adaptive manner according to distribution difference between the samples in the source-field set and the target-field set, using the better noise interference coefficient to interfere with feature space, using a stacking edge noise reduction encoder method to construct new feature space, and constructing a classifier. According to the method, relationships between potential featuresbetween fields can be better mined, field difference can be reduced, and thus correctness of classification can be improved.

Description

technical field [0001] The invention relates to a cross-domain text classification method based on an adaptive noise reduction encoder, and classifies network text data information, more specifically, cross-domain classification for network text data information in different fields and different data distributions . Background technique [0002] In recent years, with the rapid rise of online social platforms such as blogs, WeChat, and Weibo, a large amount of text information has been generated on the Internet. These massive data often contain huge potential commercial value. Information can be used to improve or upgrade products in a targeted manner, so as to meet consumer needs and increase market competitiveness; It will be more favored by consumers. In view of this, research in related fields such as text classification has extremely important value and significance. [0003] However, because the data in the network is affected by multiple factors such as users and ti...

AI Technical Summary

Problems solved by technology

Cross-domain classification tasks need to process features of multiple different domains at the same time, which further enlarges the feature space and further exacerbates the high-dimensional and sparse nature of text data, which makes it difficult to select meaningful features and provides a common way for learning. Feature spaces for cross-domain classification pose challenges

[0007] Second, although the edge denoising encoder can learn a relatively robust feature space in cross-domain classification tasks, however, its learning results are sensitive to noise coefficients

Images