Context feature fused aspect-level sentiment classification method and device

Abstract

The invention discloses a context feature fused aspect-level sentiment classification method and device. The method comprises the steps of carrying out weak correlation aspect word static shielding and word segmentation processing on the aspect-level sentiment analysis text to be predicted; obtaining a first global context, the first BERT embedded layer and the second BERT embedded layer respectively extract a first global context feature and a first local context feature, extracting a mixed local context feature by the local feature learning layer, and extracting a second global context feature by the MHSA layer; processing a fusion feature obtained after the second global context feature and the mixed local context feature are fused by the interactive learning layer; and finally, outputting an emotion polarity result by the output layer, so that the technical problem of low accuracy of emotion polarity prediction due to neglect of local contexts of aspect words, strong semantic association characteristics of the aspect words and interference of weakly related aspect words on emotion classification in the prior art is solved.

Description

technical field [0001] The present application relates to the technical field of sentiment classification, in particular to a method and device for aspect-level sentiment classification by fusing contextual features. Background technique [0002] Aspect-level sentiment classification refers to predicting the sentiment polarity of given aspect words in comment text or other texts. Unlike sentence-level sentiment classification, aspect-level sentiment classification is a fine-grained sentiment classification task. [0003] At present, aspect-level sentiment classification methods are mainly based on deep learning, which focuses on the field of recurrent neural network and convolutional neural network. Like most natural language processing tasks, aspect-level sentiment classification also employs a sequence-to-sequence encoding model. Existing aspect-level sentiment classification methods only consider the semantic association between aspect words and sentence-level context, b...

AI Technical Summary

Problems solved by technology

[0004] This application provides an aspect-level sentiment classification method and device that integrates context features, which is used to solve the strong semantic correlation characteristics of the local context of the aspect words and the strong semantic correlation characteristics of the aspect words and weakly related or irrelevant aspect words in the prior art. The interference of classification makes the emotional polarity prediction have a technical problem of low accuracy

Images