Aspect-level sentiment analysis method based on convolutional neural network

Abstract

The invention discloses an aspect-level sentiment analysis method based on a convolutional neural network. The method comprises the steps of establishing an aspect-level sentiment analysis model; constructing a relative position matrix of aspect-level information in the text through the model; fusing the text codes with the text codes; extracting a part of incidence matrixes corresponding to aspect-level information in the incidence matrixes; wherein the aspect-level emotion analysis model comprises a word embedding module, a relative position coding module, an aspect-level attention module and an sentiment classification module, and the aspect-level sentiment analysis model comprises a word embedding module, a relative position coding module, an aspect-level attention module and a sentiment classification module. According to the method, the sentiment of the aspect-level information in the text is modeled, so that the aspect-level emotion of the text is efficiently and accurately analyzed.

Description

technical field [0001] The invention relates to the fields of natural language processing and artificial intelligence, in particular to an aspect-level sentiment analysis method based on a convolutional neural network. Background technique [0002] With the rapid development and wide application of Internet technology, more and more people tend to express their opinions, express their emotions or expound their opinions through the Internet. The vigorous development of various new network platforms such as social networking, e-commerce and self-media has led to geometric growth of Internet information. How to analyze and mine these information, identify their emotional tendencies, predict their opinions, predict their evolution over time, and help users efficiently, accurately and comprehensively obtain the content that people care about from massive text information , and organize and process the information, so that users can get clear and simple and intuitive information ...

AI Technical Summary

Problems solved by technology

For this reason, the patent CN 109472031 A discloses an aspect-level emotion classification model and method based on dual memory attention, which constructs an encoder and decoder through a GRU recurrent neural network, and then uses a Softmax classifier to classify text, but the GRU The output of each step of the cyclic neural network includes the output of the previous step, so the model cannot be processed in parallel very well, and the efficiency is low

In the currently published patents and literature on aspect-level sentiment analysis, there are not many considerations about the relative position information of aspect-level information in the text, but in semantics, it is generally believed that in a piece of text, the closer the word is to the The greater the influence, the relative information of the position is very important, especially when the text contains multiple aspects of information, the emotions of the aspects of the information are easy to interfere with each other, and the relative position information is more important at this time. Therefore, An aspect-level sentiment analysis method to solve such problems is urgently needed in the field

Images