Food Public Opinion Prevention and Control Model Based on Sentiment Analysis

Area of Science:

• Food science and technology
• Computational social science
• Artificial intelligence

Background:

• Food public opinion significantly impacts food safety and consumer trust due to its rapid spread and negativity.
• Existing methods lack comprehensive strategies for managing the dynamic nature of food public opinion.

Purpose of the Study:

• To develop and validate a Food Public Opinion Prevention and Control (FPOPC) model using deep learning and recommendation algorithms.
• To enhance food safety and consumer trust through proactive management of public sentiment.

Main Methods:

• Developed a comprehensive FPOPC framework covering all stages of public opinion development.
• Utilized a Stacked Autoencoder (SAE) for sentiment prediction in food news based on user comments.
• Implemented a collaborative filtering recommendation mechanism and an enhanced Bloom filter with HDFS for rapid opinion management.

Main Results:

• The FPOPC model accurately predicts and controls the development of food public opinion.
• Sentiment analysis of food news was effectively quantified and utilized for personalized recommendations.
• Experimental validation confirmed the model's efficacy in managing food public sentiment across the supply chain.

Conclusions:

• The FPOPC model provides a robust solution for predicting and controlling food public opinion.
• This approach offers valuable tools for regulatory agencies to safeguard food safety and consumer confidence.