Food and medicine public opinion analysis method

Abstract

The invention provides a food and medicine public opinion analysis method. The method comprises the following steps: collecting a web text, and preprocessing the web text; performing sentence segmentation processing on the preprocessed web text, and removing repeated comment sentences in the web text; performing word segmentation processing on each comment sentence after sentence segmentation processing, and extracting food and medicine feature words in the text; calculating similarities between the food and medicine feature words, and if any two similarities are smaller than a set threshold value, removing one feature word; constructing a classification dictionary, and dividing the food and medicine feature words in each comment sentence into a corresponding dictionary category; constructing an emotion dictionary, and recognizing food and medicine feature words, emotion words, degree words and negative words from the comment sentences; determining a basic sentiment value of the sentiment word, a weight value of the degree word and a weight value of the negative word; according to the method, the food and medicine emotional tendency value calculation model is constructed, and the public opinion tendency of the collected web text is determined according to the emotional tendency value calculation model, so that the evaluation published by the user on the network can be accurately processed, and the accurate emotional tendency value is obtained; therefore, an accurate public opinion reference basis can be provided for food and medicine manufacturers and quality supervision departments, and accurate data support is provided for quality feedback and market supervision measure formulation.

Description

technical field [0001] The invention relates to a public opinion analysis method, in particular to a food and medicine public opinion analysis method. Background technique [0002] Food and medicine are two major themes related to people's livelihood. There are countless types of food and medicine on the market. However, users' evaluation of a certain brand of food or medicine after use is related to the subsequent evaluation of food and medicine. The implementation of behaviors such as quality feedback and market supervision. [0003] With the development of network technology and computer technology, users' evaluation of food or medicine is often stated through network methods such as Weibo and Post Bar. In the prior art, the analysis of network texts about food and medicine public opinion is based on sentiment analysis. That is, through the processing of network texts and the calculation of sentiment values, etc., however, the existing sentiment analysis methods have low...

AI Technical Summary

Problems solved by technology

[0003] With the development of network technology and computer technology, users' evaluation of food or medicine is often stated through network methods such as Weibo and Post Bar. In the prior art, the analysis of network texts about food and medicine public opinion is based on sentiment analysis. That is, through the processing of network texts, emotional value calculations, etc., however, the existing sentiment analysis methods have low accuracy, so they cannot accurately grasp the tendency of user evaluation.

Images