Specific target emotion classification method based on deep neural network

Abstract

The invention provides a specific target emotion classification method based on a deep neural network. The invention belongs to the field of text sentiment classification of natural language processing. The method comprises the following steps: firstly, performing operations of Chinese word segmentation, stop word removal and punctuation removal on a data set; training the processed corpus by adopting a word2vec algorithm to obtain a corresponding word vector; inputting the training set into a long-term and short-term memory network model structure based on a target attention mechanism; duringthe process of realizing the attention weight training process, embedding the specific target and the specific aspect and representing the specific target by using the weighted summation embedded inthe specific aspect, so that the model gives more correct attention to the specific target and the specific aspect, the real semantics of the target is better captured, and finally, the emotion classification accuracy of the specific target is improved.

Description

technical field [0001] The invention relates to comment text sentiment classification, in particular to a specific target sentiment classification method based on a deep neural network, which belongs to the technical field of natural language processing. Background technique [0002] Sentiment analysis methods mainly include rule-based methods, machine learning-based methods and deep neural network-based methods. Rule-based methods usually need to construct an emotional dictionary or emotional collocation template, and then calculate the emotional tendency of the text by comparing the emotional words or fixed collocations contained in the comment text, but building a relatively complete emotional dictionary or related collocation rules is now The main problem exists. The method based on machine learning mainly performs feature extraction and modeling on the training corpus with labels, so as to automatically realize the judgment of emotional polarity with machine learning a...

AI Technical Summary

Problems solved by technology

However, most of the neural network-based text sentiment classification models do not get the correct attention to the emotion of a specific aspect of a specific target, and the classification effect is relatively poor.

Images