Emotion tendency classification method and device based on user comments and storage medium

Abstract

The invention discloses an emotion tendency classification method and device based on user comments and a storage medium, and the method comprises the steps: carrying out the preprocessing of a user comment statement, and obtaining a lexical item; obtaining a word vector matrix; obtaining previous semantic features and next semantic features of each lexical item according to a bidirectional gatingcycle unit, and splicing the previous semantic features, the next semantic features and lexical vectors of the lexical items; extracting a feature map with a fixed dimension from the spliced word vectors by using a convolutional neural network; classifying the emotional tendency according to the feature map; and displaying the merchant on a map according to the classification result. The method has the advantages that the problem that a large number of rules and dictionaries need to be manually summarized in a rule-based method and the problem that semantics and word orders are lost on the basis of a word bag model are solved by combining the bidirectional gating circulation unit and the convolutional neural network; and the large-scale comment data is labeled according to the category ofthe emotional tendency, so that the problem of time and labor waste of manual labeling is solved.

Description

technical field [0001] The present application relates to the technical field of natural language processing, and in particular to a method, device and storage medium for classifying sentiment tendency based on user comments. Background technique [0002] At present, based on the user's comments on merchants, the user's emotions are analyzed to discover high-quality merchants and update the high-quality merchants to the map. User sentiment analysis is usually achieved through rule-based methods and machine learning-based classification methods. Rule-based methods usually need to rely on a large number of corpus resources, such as dictionaries and corresponding rules that are artificially summarized, and use dictionaries and rules to match the corresponding terms in the document to obtain text sentiment. This method needs to summarize a large number of rules based on human experience, and its performance will be greatly affected by the output results of word segmentation, pa...

AI Technical Summary

Problems solved by technology

Solve the problem that a large number of rules and dictionaries need to be manually summarized and will be affected by errors in the underlying analysis tools; solve the problem of semantic and word order loss based on the bag-of-words model; solve the time-consuming and labor-intensive problems of manual labeling

Images