E-commerce comment sentiment analysis system based on deep learning

Abstract

The invention relates to the technical field of deep learning, and discloses an e-commerce comment sentiment analysis system based on deep learning. The system comprises a comment extraction module, a polarity classification module, a deep learning module, a word vector matching module, an emotion attribute word extraction module, an emotion attribute word fusion module, a word meaning and word order loss correction module and an attribute emotion label extraction module. A word vector language representation model extracted through the word vector matching module and the emotion attribute word extraction module and emotion attribute words are calculated and fused through the emotion attribute word fusion module to form a new attribute emotion initialization tag, the word meaning and word order loss correction module reads the attribute emotion initialization tag, the correctness of the word order and the word meaning of the attribute emotion initialization tag is determined through logic calculation, and if the word meaning or the word order is determined to have a logic error, the word meaning and the word order are revised, so that the emotion expression of the emotion word is increased, and the expression form and the expression effect are further increased.

Description

technical field [0001] The invention relates to the technical field of deep learning, in particular to an e-commerce comment sentiment analysis system based on deep learning. Background technique [0002] Deep learning is a new research direction in the field of machine learning, which is introduced into machine learning to make it go further with the original goal - artificial intelligence. Deep learning is to learn the internal laws and representation levels of sample data. The information obtained in the learning process is of great help to the interpretation of data such as text, images and sounds, which has made great progress in artificial intelligence related technologies. [0003] With the development of the Internet and e-commerce, the product review mode of e-commerce has also changed. In order to facilitate the increase of the expression form of the product, attribute tags are usually added to the product. The existing attribute tags are usually generated by direc...

AI Technical Summary

Problems solved by technology

The existing label description is simple, the semantics are monotonous, and the description accuracy of the attribute label is revealed due to the direct extraction and simplification of the product theory words, and the existing attribute word extraction usually only includes product words and functional words. The lack of description of emotional words leads to insufficient expression of product attribute labels

Images