Sentiment analysis method, system and device based on weight dictionary and storage medium

Abstract

The invention relates to the technical field of artificial intelligence, and provides a sentiment analysis method, system and device based on a weight dictionary and a storage medium. The method comprises the steps that a Chinese sentiment analysis data set is collected, word segmentation and weight calculation are conducted on the Chinese sentiment analysis data set, a sentiment type weight dictionary is constructed, and the Chinese sentiment analysis data set carries positive or negative annotation data; the method also includes obtaining a plurality of pieces of text data, and performing positive and negative judgment and labeling on each piece of text by adopting the sentiment type weight dictionary to obtain labeling information; carrying out vector conversion on the annotation information of each text, and extracting to obtain a feature vector of the text; and performing parameter adjustment training on a pre-training model according to the feature vector, outputting a trained model, and performing sentiment analysis by adopting the trained model. According to the invention, the problems of low automation degree, complex operation, low speed and low accuracy in the prior art can be solved.

Description

technical field [0001] The invention relates to the technical field of artificial intelligence, in particular to a weight dictionary-based sentiment analysis method, system, device and storage medium. Background technique [0002] The problem of text sentiment analysis is one of the most important tasks in the field of artificial intelligence NLP. For humans, it is easy for people to understand the author's emotion through the text, but for the machine, how to make the machine recognize the emotion contained in the text is difficult. [0003] Sentiment analysis has a huge impact in real life. For example, consumers can decide whether to buy a product based on other consumers’ comments on the product; catering businesses can use consumer evaluation information to assess whether the service quality is not good enough. Food quality and taste do not meet the requirements of the public, etc. [0004] Traditional sentiment classification systems often require a large amount of m...

AI Technical Summary

Problems solved by technology

[0005] Based on the above problems, the object of the present invention is to provide a sentiment analysis method, system, device and storage medium based on a weight dictionary, to solve the problem of low automation of sentiment analysis in the prior art, complicated operation, slow speed and high accuracy. not high problem

Images